The Ratio Club: A Hub of British Cybernetics1
Phil Husbands and Owen Holland
1. Introduction
Writing in his journal on the 20th September 1949, W. Ross Ashby noted that six days
earlier he’d attended a meeting at the National Hospital for Nervous Diseases,
London. He comments, “We have formed a cybernetics group for discussion – no
professors and only young people allowed in. How I got in I don’t know, unless my
chronically juvenile appearance is at last proving advantageous. We intend just to talk
until we can reach some understanding.” (Ashby 1949a). He was referring to the
inaugural meeting of what would shortly become the Ratio Club, a group of
outstanding scientists who at that time formed much of the core of what can be
loosely called the British cybernetics movement. The club usually gathered in a
basement room below nurses’ accommodation in the National Hospital, where, after a
meal and sufficient beer to lubricate the vocal chords, participants would listen to a
speaker or two before becoming embroiled in open discussion. The club was founded
and organized by John Bates, a neurologist at the National Hospital. The other twenty
carefully selected members were a mixed group of mainly young neurobiologists,
engineers, mathematicians and physicists.
A few months before the club started meeting, Wiener’s landmark Cybernetics:
Control and Communication in the Animal and Machine (Wiener 1948) had been
published. This certainly helped to spark widespread interest in the new field, as did
Shannon’s seminal papers on information theory (Shannon and Weaver 1949), and
these probably acted as a spur to the formation of the club. However, as we shall see
later, the first of the official membership criteria of the club was that only “… those
who had Wiener’s ideas before Wiener’s book appeared … ” (Bates 1949a) could
join. This was no amateur cybernetics appreciation society; many members had
already been active for years in developing the new ways of thinking about behaviour
generating mechanisms and information processing in brains and machines that were
now being pulled together under the term coined by Wiener. Indeed, the links and
mutual influences that existed between the American and British pioneers in this area
ran much deeper than is often portrayed. There was also a very strong independent
British tradition in the area that had developed considerable momentum during World
War II. It was from this tradition that most club members were drawn.
The other official membership criterion reflected the often strongly hierarchical nature
of professional relationships at that time. In order to avoid restricting discussion and
debate, Bates introduced the ‘no professors’ rule alluded to by Ashby. If any members
should be promoted to that level, they were supposed to resign. Bates was determined
to keep things as informal as possible; conventional scientific manners were to be
eschewed in favour of relaxed and unfettered argument. There also appears to have
1
Published in P. Husbands, O. Holland, M. Wheeler (eds) The Mechanical Mind in History, MIT
Press, 91-148, 2008.
1
been two further, unofficial, criteria for being invited to join. First, members had to be
as smart as hell. Second, they had to be able to contribute in an interesting way to the
cut and thrust of debate, or, to use the parlance of the day, be good value. This was a
true band of Young Turks. In the atmosphere of enormous energy and optimism that
pervaded post-war Britain as it began to rebuild, they were hungry to push science in
new and important directions. The club met regularly from 1949-1955, with one final
reunion meeting in 1958. It is of course no coincidence that this period parallels the
rise of the influence of cybernetics, a rise in which several members played a major
role.
There are two things that make the club extraordinary from a historical perspective.
The first is the fact that many of its members went on to become extremely prominent
scientists. The second is the important influence the club meetings, particularly the
earlier ones, had on the development of the scientific contributions many of that
remarkable group would later make. The club membership undoubtedly made up the
most intellectually powerful and influential cybernetics grouping in the UK, but to
date very little has been written about it – there are brief mentions in some histories of
AI and cognitive science (e.g. Fleck 1982, Boden 2006, and D. Clark (2003) has a
chapter on it in his PhD thesis, based on papers from the John Bates archive). This
article is intended to help fill that gap. It is based on extensive research in a number of
archives, interviews with surviving members of the club and access to some
member’s papers and records.
After introducing the membership in the next section, the birth of the club is described
in some detail. The club’s known meetings are then listed and discussed along with its
scope and modus operandi. Following this some of the major themes and
preoccupations of the club are described in more detail. The interdisciplinary nature
of the intellectual focus of the group is highlighted before discussing the legacy of the
club. Because so many rich threads run through the club and the lives and work of its
members, this chapter can only act as an introduction. A fuller treatment of all these
topics can be found in Husbands and Holland (forthcoming).
2. The members
Before embarking on a description of the founding of the club, it is useful at this point
to sketch out some brief details of its membership as that will help to give a sense of
the historical importance of the group.
The definitive list of twenty-one members, with very brief details of expertise and
achievements, is given below. Of course these summaries are far too short to do
justice to the careers of these scientists. They are merely intended to illustrate the
range of expertise in the club and to give a flavour of the calibre of members.
W. Ross Ashby (1903-1972), trained in medicine and psychiatry, is regarded as one
of the most influential pioneers of cybernetics and systems science. Author of the
classic books Design for a Brain (Ashby 1952a) and An Introduction to Cybernetics
(Ashby 1958), some of his key ideas have recently experienced something of a
renaissance in various areas of science including Artificial Life and modern AI. At the
inception of the club he was director of research at Barnwood House Hospital,
2
Gloucester. He subsequently became a professor in the Department of Biophysics and
Electrical Engineering, University of Illinois.
Horace Barlow FRS (1921- ), a great-grandson of Charles Darwin, is an enormously
influential neuroscientist, particularly in the field of vision, and was one of the
pioneers of using information theoretic ideas to understand neural mechanisms
(Barlow 1953, 1959, 1961), a direct consequence of his involvement in the Ratio
Club. When the club started he was a PhD student in Lord Adrian’s lab at the
department of physiology, Cambridge University. He later became Royal Society
Research Professor of Physiology at Cambridge University.
John Bates (1918-1993) had a distinguished career in the neurological research unit
at The National Hospital for Nervous Diseases, London. He studied human EEG in
research into voluntary movement and became the chief electroencephalographer at
the hospital. The Club was his idea and he ran it with quiet efficiency and unstinting
enthusiasm.
George Dawson (1911-1983) was a clinical neurologist at the National Hospital,
Queen’s square. At the time of the Ratio Club he was a world leader in using EEG
recordings in a clinical setting. He was a specialist in ways of averaging over many
readings which allowed him to gather much cleaner signals than was possible by more
conventional methods (Dawson 1954). He became Professor of Physiology at UCL.
Thomas Gold FRS (1920-2004) was one of the great astrophysicists of the 20th
century, being a co-author, with Bondi and Hoyle, of the steady state theory of the
universe and having given the first explanation of pulsars, among countless other
contributions. However, he had no time for disciplinary boundaries and at the time of
the Ratio Club he was working in Cambridge University Zoology Department on a
radical positive feedback theory of the working of the inner ear (Gold 1948) – a
theory that was, typically for him, decades ahead of its time. He went on to become
Professor of Astronomy at Harvard University and then at Cornell University.
I.J. (Jack) Good (1916- 2009) was recruited into the UK top secret code cracking
operation at Bletchley Park during the second world war, where he worked as the
main statistician under Alan Turing and Max Newman. Later he became a very
prominent mathematician, making important contributions in Bayesian methods and
early AI. During the Ratio years he worked for British Intelligence. Subsequently he
became Professor of Statistics at Virginia Polytechnic Institute.
W.E. Hick (1912-1974) was a pioneer of information theoretic thinking in
psychology. He is the source of the still widely quoted Hick’s law which states that
the time taken to make a decision is proportion to the log of the number of alternatives
(Hick 1952). During the Ratio years he worked in the Psychology laboratory at
Cambridge University. He went on to become a distinguished psychologist.
Victor Little (1920-1976) was a physicist at Bedford College, London, who worked
in acoustics and optics before moving on to laser development.
Donald Mackay (1922-1987), trained as a physicist, was a very highly regarded
pioneer of early machine intelligence and of neuropsychology. He was also the
3
leading scientific apologist for Christianity of his day. At the birth of the club he was
working on a PhD in the Physics department of King’s College, London. He later
became a professor at Keele University where he founded the Department of
Communication and Neuroscience.
Turner McLardy (1913-1988) became an international figure in the field of clinical
psychiatry. He emigrated to the USA in the late 1950s to develop therapeutic
techniques centred around planned environments and communities. Later he became a
pioneer of understanding the role of zinc in alcoholism and schizophrenia. At the
inception of the club he worked at Maudsley Hospital, London.
Pat Merton FRS (1921-2000) was a neurophysiologist who did pioneering work on
control theoretic understandings of the action of muscles (Merton 1953). Later he
carried out a great deal of important early research in magnetic stimulation of the
cortex for which he is justly celebrated (Merton and Morton 1980). During the Ratio
years he worked in the neurological research unit at the National Hospital. He later
became Professor of Human Physiology at Cambridge University.
John Pringle FRS (1912-1982) was one of the leading invertebrate neurobiologists of
his day. He was the first scientist to get recordings from single neurons in insects,
something that had previously been thought to be practically impossible (Pringle
1938). He did much important work in proprioception in insects, insect flight and
invertebrate muscle systems. At the birth of the club he worked in the Zoological
laboratory, Cambridge University. He subsequently became Professor of Zoology at
Oxford University.
William Rushton FRS (1901-1980) is regarded as one of the great figures in 20th
century vision science. He made enormous contributions to understanding the
mechanisms of colour vision, including being the first to demonstrate the deficiencies
that lead to colour blindness (Rushton 1955). Earlier he did pioneering work on the
quantitative analysis of factors involved in the electrical excitation of nerve cells,
helping to lay the foundations for the framework that dominates theoretical
neuroscience today (e.g. Rushton 1935). He worked at Cambridge University
throughout his career where he became Professor of Visual Physiology.
Harold Shipton (1920- 2007) worked with Grey Walter on the development of EEG
technology at the Burden Neurological Institute, Bristol. He was the electronics
wizard who was able to turn many of Walter’s less than precise designs into working
realities. Later he became a professor at The University of Washington at St. Louis,
where he worked on biomedical applications. At the time of the Ratio meetings, his
father-in-law, Clement Attlee, was prime minister of Great Britain.
D.A. Sholl (1903-1960) did classic research on describing and classifying neuron
morphologies and growth patterns, introducing the use of rigorous statistical
approaches (Sholl 1956). Most of the classification techniques in use today are based
on his work. He also published highly influential papers on the structure and function
of the visual cortex. He worked in the Anatomy department of University College,
London where he became Reader in Anatomy before dying young.
4
Eliot Slater (1904-1983) was one of the most eminent British psychiatrists of the
twentieth century. He helped to pioneer the use of properly grounded statistical
methods in clinical psychiatry. Slater’s work with Rudin on the genetic origins of
schizophrenia, carried out in Munich in the 1930s, still underpins all respectable
Anglo-American work in psychiatric genetics, a field to which Slater made many
important contributions (Slater et al. 1971). He worked at the National Hospital for
Nervous diseases, London.
Alan Turing FRS (1912-1954) is universally regarded as one of the fathers of both
computer science and artificial intelligence. Many regard him as one of the key
figures in twentieth century science and technology. He also anticipated some of the
central ideas and methodologies of Artificial Life and Nouvelle AI by half a century –
for instance, he proposed artificial evolutionary approaches to AI in the late 1940s
(Turing 1950) and published work on reaction-diffusion models of the chemical
origins of biological form in 1952 (Turing 1952). At the inception of the club he was
working at Manchester University, where he was part of a team that had recently
developed the world’s first stored-program digital computer.
Albert Uttley (1906-1985) did important research in radar, automatic tracking and
early computing during WWII. Later he became head of the pioneering Autonomics
Division at the National Physical Laboratory in London where he did research on
machine intelligence and brain modeling. However, he also became well know as a
neuropsychologist, having made several important contributions to the field (Uttley
1979). At the birth of the club he worked at TRE, Malvern, the main British military
telecommunications research establishment. Later he became Professor of Psychology
at Sussex University.
W. Grey Walter (1910-1977) was a world leader in EEG research. He discovered
theta and delta brain waves and, with Shipton, developed the first EEG brain
topography machine (Walter and Shipton 1951). At the time of the Ratio Club he was
at the Burden Neurological Institute, Bristol, where, alongside his EEG research, he
developed the first ever autonomous mobile robots, referred to as tortoises, which
were controlled by analogue electronic nervous systems (Walter 1950a). This was the
first explicit use of mobile robots as a tool to study ideas about brain function, a style
of research that has become very popular in recent times.
John Westcott FRS (1920- ) made many very distinguished contributions to control
engineering, including some of the earliest work on control under noisy conditions.
He also worked on applications of control theory to economics which resulted in his
team developing various models used by the UK Treasury. At the inception of the
club he was doing a PhD in the department of Electrical Engineering, Imperial
College, London, having just returned from a year in Wiener’s lab at MIT. He later
became Professor of Control Systems at Imperial College.
Philip M. Woodward (1919- ) is a mathematician who made important contributions
to information theory, particularly with reference to radar, and to early computing.
His gift for clear concise explanations can be seen in his elegant and influential 1953
book on information theory (Woodward 1953). He worked at TRE, Malvern
throughout his entire distinguished career (one of the buildings of the present day
successor to TRE is named after him). In retirement Woodward has come to be
5
regarded as one of the world’s greatest designers and builders of mechanical clocks
(Woodward 1995).
Bates’ own copy of his typed club membership list of 1 st January 1952 has many
hand-written corrections and annotations (Bates 1952a). Among these, immediately
under the main list of members, are the following letters, arranged in a neat column:
‘Mc’, ‘P’, ‘S’ and then a symbol that may be a ‘U’ or possibly a ‘W’. If we assume it
is a ‘W’, then a possible, admittedly highly speculative, interpretation of these letters
is: McCulloch, Pitts, Shannon, Wiener. The first three of these great American
cyberneticists attended club meetings – McCulloch appears to have taken part
whenever travel to Britain allowed. Wiener was invited and intended to come on at
least one occasion but travel difficulties and health problems appear to have got in the
way. The ‘W’, if that’s what it is, could also refer to Weaver, co-author with Shannon
of seminal information theory papers and someone who was also well known to the
club. Of course the letters may not refer to American cyberneticists at all – they may
be something more prosaic such as the initials of members who owed subs – but it is
just possible that Bates regarded them as honorary members.
It is clear from the membership listed above that the centre of gravity of the club was
in the brain sciences. Indeed the initial impetus for starting the club came from a
neurologist (Bates) who believed that emerging cybernetic ideas and ways of thinking
could be very important tools in developing new insights into the operation of the
nervous system. Many members had a strong interest in developing ‘brain-like’
devices, either as a way of formalizing and exploring theories about biological brains,
or as a pioneering effort in creating machine intelligence, or both. Hence meeting
tended to centre around issues relating to natural and artificial intelligence and the
processes underlying the generation of adaptive behaviour – in short, the
mechanisation of mind. Topics from engineering and mathematics were usually
framed in terms of their potential to shed light on these issues. This scope is
somewhat different to that which had emerged in America, where a group of
mathematicians and engineers (Wiener, von Neumann, Bigelow, Shannon, Pitts) and
brain scientists (Lorente de No, Rosenblueth, McCulloch) had formed an earlier
group similar in spirit to the Ratio Club, although smaller and with a centre of gravity
further towards the mathematical end of the spectrum. Their influence soon spread,
via Frank, Mead, Bateson and others, into the social sciences, thereby creating a much
wider enterprise (Heims 1991). This difference in scope helps to account for the
distinct flavour of the British scene in the late 1940s and for its subsequent influences.
3. Genesis of the club
3.1 Founding
The idea of forming a cybernetics dining club took root in John Bates’ mind in July
1949. He discussed the idea with a small number of colleagues at a Cambridge
symposium on ‘Animal Behaviour Mechanisms’, a very cybernetics friendly topic,
organised by the Society for Experimental Biology and held from the 18th to 22nd of
the month. Shortly after returning to London from the meeting, he wrote the following
letter to Grey Walter in which the club is formally proposed:
6
“National Hospital
27th July 1949
Dear Grey,
I have been having a lot of ‘Cybernetic’ discussions during the past
few weeks here and in Cambridge during a Symposium on Animal
Behaviour Mechanisms, and it is quite clear that there is a need for the
creation of an environment in which these subjects can be discussed
freely. It seems that the essentials are a closed and limited membership
and a post-prandial situation, in fact a dining-club in which
conventional scientific criteria are eschewed. I know personally about
15 people who had Wiener’s ideas before Wiener’s book appeared and
who are more or less concerned with them in their present work and
who I think would come. The idea would be to hire a room where we
could start with a simple meal and thence turn in our easy chairs
towards a blackboard where someone would open a discussion. We
might need a domestic rule to limit the opener to an essentially
unprepared dissertation and another to limit the discussion at some
point to this stratosphere, but in essence the gathering should evolve in
its own way.
Beside yourself, Ashby and Shipton, and Dawson and Merton from
here, I suggest the following:Mackay – computing machines, Kings. Coll. Strand
Barlow – sensory physiologist – Adrian’s lab.
Hick – Psychological lab. Cambridge
Scholl – statistical neurohistologist – UniversityCollege, Anatomy
Lab.
Uttley – ex. Psychologist, radar etc TRE
Gold – ex radar zoologists at Cambridge
Pringle – ex radar zoologists at Cambridge
I could suggest others but this makes 13, I would suggest a few more
non neurophysiologists communications or servo folk of the right sort
to complete the party but those I know well are a little too senior and
serious for the sort of gathering I have in mind.
We might meet say once a quarter and limit the inclusive cost to 5/less drinks. Have you any reaction? I have approached all the above
list save Uttley so far, and they support the general idea.
Yours sincerely,
JAV Bates”
(J.Bates 1949a)
The suggested names were mainly friends and associates of Bates, known through
various social networks relating to his research, who he regarded as being ‘of the right
sort’. One or two were suggested by immediate colleagues, for instance Merton put
forward his friend Barlow.
7
Walter replied by return of post enthusiastically welcoming the idea and suggesting
that the first meeting should coincide with his friend Warren McCulloch’s visit to
England in September. Mackay furnished Bates with an important additional
‘communications or servo’ contact by introducing him to John Westcott who was
finishing off his PhD at Imperial College, having spent the previous year in Wiener’s
lab at MIT as a guest of the institution. Westcott’s close association with Wiener
seems to have led Bates to soften his ‘had Wiener’s ideas before Wiener’s book
appeared’ line in his invitation to him.
“National Hospital
3rd August
Dear Mr. Westcott,
I have heard from Mackay that you might be interested in a dining-club
that I am forming to talk ‘Cybernetics’ occasionally with beer and full
bellies. My idea was to have a strictly limited membership between 15
and 20, half primarily physiologists and psychologists though with
‘electrical leanings’ and half primarily communication theory and
electrical folk though with biological interests and all who I know to
have been thinking ‘Cybernetics’ before Weiner’s book appeared. I
know you have all the right qualifications and we would much like you
to join. The idea is to meet somewhere from 7.00pm-10.00pm at a cost
of about 5/- less drinks.
The second point is whether we could make McCulloch’s visit in
September the occasion for a first meeting. This was raised by Mackay
who mentioned that you had got in touch with him already with a view
to some informal talk. It has also been raised by Grey Walter from
Bristol who knows him too. What do you feel? Could we get
McCulloch along to an inaugural dinner after his talk for you? Could
you anyway manage to get along here for lunch one day soon, we have
an excellent canteen and we could talk it over?
Your sincerely
JAV Bates” (J.Bates 1949b)
Westcott was as enthusiastic as Walter. Bates wrote a succession of individual
invitations to those on his list as well as to Little, who was suggested by Mackay, and
Turner McLardy, a psychiatrist with a keen interest in cybernetics who was a friend of
McCulloch and appears to have been hosting his immanent stay in London. The letter
to Hick was typical, including the following exuberant passage: “The idea of a
‘Cybernetic’ dining club, which I mentioned to you in Cambridge, has caught fire in
an atomic manner and we already have half a dozen biologists and engineers, all to
my knowledge possessed of Wiener’s notions before his book appeared and including
two particularly rare birds: Mackay and Westcott who were in Wiener’s lab for a year
during the war…” (Bates 1949c). Bates didn’t quite have his facts straight; Westcott’s
time with Wiener was after the war and at this stage Mackay hadn’t begun his
8
collaborations with MIT, but the implication was right – that Westcott and Mackay
were both familiar with the mathematical and technical details of Wiener’s oevre.
All invitees accepted membership of the club. In their replies a number made general
suggestions about membership – Barlow suggested considering the addition of a few
more “cautiously selected psychologists” (Barlow 1949), and Pringle thought it
would be a good idea to “add a mathematician to keep everyone in check and stop the
discussion becoming too vague.” (Pringle 1949)
During August Bates secured a room at the National Hospital that could be used for
regular meetings. With Eliot Slater, a senior member of staff at the hospital, on-board,
he was able to arrange provision of beer and food for club evenings. With a venue, a
rough format and an initial membership list, the enterprise was starting to come into
focus. The following letter from Mackay to Bates, hand written in a wild scrawl,
shows that these two were starting to think about names and even emblems:
“1st September 49
Dear Bates,
I’m afraid I’ve had few fresh ideas on the subject of our proposed club;
but here are an odd suggestion or two that arose in my mind.
I wondered (a) if we might adopt a Great Name associated with the
subject and call it e.g. the Babbage Club or the Leibniz Club or the
Boole Club, or the Maxwell Club – names to be suggested by all, and
one selected by vote or c’ttee (Nyquist might be another). Alternatively
(b) could we choose a familiar symbol of feedback theory, such as
beta, and call it the Beta Club or such like? Other miscellaneous
possibilities are the MR Club (machina ratiocinatrix !) and plenty of
other initials, or simply the “49” Club.
On emblems I’ve had no inspirations. I use but little beer myself and
it’s conceivable we might even have t-t members. But beer mugs can
after all be used for other liquids and I can’t think of anything better
than your suggestion…
Yours,
Donald Mackay” (Mackay 1949)
Here we see Mackay sowing the seed for the name Ratio, which was adopted after the
first meeting. Machina Ratiocinatrix is Latin for reasoning machine, a term used by
Wiener in the introduction to Cybernetics in reference to calculus ratiocinator, a
calculating machine constructed by Leibniz (Wiener 1948, p.12). Ratiocination is an
old-fashioned word for reasoning or thinking, introduced by Thomas Aquinas to
distinguish human reasoning from the supposed directly God given knowledge of the
angels. After the first meeting Albert Uttley suggested using the root ratio giving its
definition as ‘computation or the faculty of mind which calculates, plans and reasons’
(Bates 1949d). He pointed out that it is also the root of rationarium, meaning a
statistical account – implicitly referring to the emerging work on statistical
mechanisms underlying biological and machine intelligence, and ratiocinatius,
meaning argumentative. Given that the name clearly came from the Latin, it seems
9
reasonable to assume that the intended pronunciation must have been ‘rat-ee-oh’. In
interviews with the authors, half the surviving club members said that this indeed is
how it was always pronounced, while the other half said it was pronounced as in the
ratio of two numbers! As Thomas Gold commented in 2002, “At that time many of
us [in the Ratio Club] were caught up in the excitement of our thoughts and ideas and
didn’t always notice the details of things like that!” (Gold 2002).
Bates’ notes for his introduction to the inaugural meeting reveal that his suggestion
was to call it the Potter Club after Humphrey Potter (Bates 1949e). Legend has it that,
as an eleven year old boy in 1713, Potter invented a way of automatically opening and
closing the valves on an early Newcomen steam engine. Until that point the valves
had to be operated by an attendant such as Potter. He decided to make his life easier
by attaching a series of cords and catches such that the action of the main beam of the
engine opened and closed the valves.
At the end of August 1949 Bates attended an EEG conference in Paris at which he
met McCulloch for the first time. There he secured him as guest speaker for the first
meeting of the club. Before describing the meetings, it will be instructive to delve a
little deeper into the origins of the club.
3.2 Origins
Of course the roots of the club go back further than the Cambridge symposium of July
1949. The second world war played an important catalytic role in developing some of
the attitudes and ideas that were crucial to the success of the Club and to the
achievements of its members. This section explores some of these roots, shedding
light on the significant British effort in what was to become known as cybernetics, as
well as pointing out pre-existing relationships in the group.
The War Effort
Many of the unconventional and multi-disciplinary ideas developed by club members
originated in secret war-time research on radar, gunnery control and the first digital
computers. In Britain there was little explicit biological research carried out as part of
the war effort, so most biologists were, following some training in electronics, drafted
into the main thrust of scientific research on communications and radar. They became
part of an army of thousands of technical ‘wizards’ who Churchill was to later
acknowledge as being vital to the allies’ victory (Churchill 1949). Although most of
the future Ratio Club biologists were naturally unconstrained and interdisciplinary
thinkers, such war work exposed many of them to more explicitly mechanistic and
mathematical ways of conceiving systems than they were used to. To these biologists
a radar set could be thought of as a kind of artificial sense organ, and they began to
see how the theoretical framework associated with it – which focused on how to best
extract information from the signal – might be applied to better understanding natural
senses such as vision. On the other side of the coin, several club members were
deeply involved in the war-time development of early computers and their use in code
cracking. This in turn brought them to ponder the possibility of building artificial
brains inspired by real ones. Other engineers and theoreticians, working on such
problems as automatic gun aiming alongside their biologist colleagues, began to see
the importance of coordinated sensing and acting in intelligent adaptive behaviour, be
it in a machine or in an animal. Many years later, in the posthumously published text
10
of his 1986 Gifford Lectures, Donald Mackay reflected on the war-time origins of his
research interests:
“...during the war I had worked on the theory of automated and electronic
computing and on the theory of information, all of which are highly relevant to
such things as automatic pilots and automatic gun direction. I found myself
grappling with problems in the design of artificial sense organs for naval gundirectors and with the principles on which electronic circuits could be used to
simulate situations in the external world so as to provide goal-directed
guidance for ships, aircraft, missiles and the like. Later in the 1940's, when I
was doing my Ph.D. work, there was much talk of the brain as a computer and
of the early digital computers that were just making the headlines as
"electronic brains." As an analogue computer man I felt strongly convinced
that the brain, whatever it was, was not a digital computer. I didn't think it was
an analogue computer either in the conventional sense. But this naturally
rubbed under my skin the question: well, if it is not either of these, what kind
of system is it? Is there any way of following through the kind of analysis that
is appropriate to there artificial automata so as to understand better the kind of
system the human brain is? That was the beginning of my slippery slope into
brain research.” (Mackay 1991)
This coalescing of biological, engineering and mathematical frameworks would
continue to great affect a few years later in the Ratio Club. As well as Mackay, of
future members at least Pringle, Gold, Westcott, Woodward, Shipton, Little, Uttley
and Walter were also involved in radar research. Hick and Bates both worked on the
related problem of visual tracking in gunnery. Uttley also worked on a range of other
problems including the development of automatic control systems, analogue computer
controlled servo mechanisms and navigation computers (for this war work he was
awarded the Simms Gold medal of the Royal Aeronautical Society). There is not
enough space in this paper to describe any of this work in detail, instead a number of
sketches are given below offering a flavour of the kinds of developments that were
undertaken and the sorts of circumstances many future members found themselves
thrust into.
Philip Woodward left Oxford University in 1941 with a degree in mathematics. As an
able bodied young man he was whisked straight into the Army where he began basic
training. However, he felt he would be much better employed at the military
Telecommunications Research Establishment (TRE) nestled in the rolling hills near
Malvern. It was here that thousands of scientists of all persuasions were struggling
with numerous seemingly impossible radar and communications problems. Within a
few days his wish was granted following a letter from his obviously persuasive father
to their local MP, Lord Beaverbrook, Minister of Supply. Leaving rifle drill far
behind, Woodward joined Henry Booker’s theoretical group to be plunged into crucial
work on antennae design and radio wave propagation. Within a few days of arriving
at TRE he was summoned to see Alec Reeves, a brilliant, highly unconventional
engineer and one of the senior staff in Woodward’s division. A few years earlier
Reeves had invented pulse code modulation, the system on which all modern digital
communication is based. He firmly believed he was in direct contact with the spirits
of various British scientific geniuses from bygone ages who through him were helping
11
in the war effort. Reeves handed Woodward a file marked ‘Top Secret’. Inside were
numerous squiggles recorded from a cathode ray tube: his task was to analyse them
and decide whether or not they came from Michael Faraday. Over the years
Woodward was to face many technical challenges almost as great as this in his work
at TRE. (Woodward 2002)
John Westcott was an engineering apprentice in the early years of the war. His job
was little more than that of a storeman, fetching and filing orders for materials to be
used in the manufacture of various military hardware. Although he didn’t have a
degree or much formal training, he was enormously frustrated by not being able to
contribute more; he was convinced that he had design talents that could really make a
difference if only he could use them (Westcott 2002). After much badgering, he
finally managed to get himself transferred to TRE where his abilities were indeed
soon recognised. He was teamed up with two other brilliant young engineers with
whom he was given complete freedom to try and design a new type of radar set to be
used by the artillery. If they were successful the device would be extremely important
– by using a significantly shorter wavelength than before it would provide a much
higher degree of accuracy. The other members of the team were the highly eccentric
Francis Farley and, on secondment from the American Signals Corps, Charles
Howard Vollum. All three were in their early twenties. At first Farley and Vollum
were always at each other’s throats with Westcott trying to keep the peace. Vollum
became incensed at the unreliability of the oscilloscopes at their disposal and swore
that after the war he’d build one that was fit for engineers to use. Despite setbacks and
failures they persevered, making use of Vollum’s supply of cigars to rope in extra
help and procure rare supplies. Somehow they managed to combine their significant
individual talents to solve the problem. This great success placed Westcott and Farley
on the road to highly distinguished scientific careers while Vollum was as good as his
word and after returning to Oregon co-founded the giant electronic instruments
company Tektronix and became a billionaire.
Like Woodward and Westcott, Thomas Gold’s route into radar research was not
direct, although his entry was rather more painful. Born into a wealthy Austrian
Jewish family, he was a student at an exclusive Swiss boarding school in the late
1930s when his father decided the political situation was becoming too dangerous to
stay in Vienna and moved the family to London. He began an engineering degree at
Cambridge University but when war broke out he was rounded up and put into an
internment camp as an enemy alien. Sleeping on the same cold concrete floor as Gold
was a young mathematician named Herman Bondi. The two struck up an immediate
friendship and began discussing the ideas that would later make them both giants of
20th Century astrophysics. Their partnership was initially short-lived because after
only a few weeks Gold was transferred to a camp in Canada. His ship survived the
brutal Atlantic crossing, although others in the convey did not, being destroyed by Uboats with the loss of many hundreds of lives. Once on Canadian soil the situation did
not improve. He found himself in a camp run by a brutally sadistic officer who made
life hell for the interns. In order to make things bearable, he claimed he was an
experienced carpenter and was put in charge of a construction gang. Ever ingenious,
he built a contraption to divert steam from an outlet pipe into a water trough to allow
his fellow interns to have a hot bath. He was severely beaten for his trouble.
Fortunately, Bondi, who had by now been rescued from another camp by senior
scientific staff who knew him at Cambridge, had been spreading word of his friend’s
12
brilliance. Gold was pulled out of internment and, like Bondi, he was assigned to
work on top secret radar research. But not before he had the great pleasure one
morning of joining with all other inmates in wild celebrations on hearing the
unexpected news that the camp commander had died of a sudden heart attack in the
night (Gold 2002).
Following a year in Princeton working with John von Neumann, Alan Turing was a
research fellow at Cambridge University when the British government, fearing war
was inevitable, recruited him into a secret codes and ciphers unit in 1938. As is well
documented (Hodges 1983), he became an enormously important figure in the
successful war-time code cracking work at Bletchley Park, and through this work was
deeply involved in the development of the very first digital computers, the theoretical
foundations for which he had set out in the late 1930s (Turing 1936). Once war broke
out, Jack Good, who had just finished a PhD in mathematics at Cambridge under the
great G.H. Hardy, was recruited into the top secret operation at Bletchley Park, where
he worked as the main statistician under Turing and Max Newman in a team that
included Donald Michie.
Most other Ratio Club members not mentioned above were medically trained and so
worked as doctors or in medical research during the war. Most of those were based in
the UK, although McLardy, who held the rank of Major, saw active service as a
medical officer and was captured and put in a succession of POW camps, at least one
of which he escaped from. He worked as a psychiatrist in Stalag 344 at Lamsdorf,
Silesia (now Lambinowice, Poland) (BBC History 2005). In early 1945 the Germans
started evacuating Lamsdorf ahead of the Russian advance. The POWs were marched
west in columns of a thousand, each column under the charge of a medical officer.
The conditions endured on these ‘death marches’ were appalling – bitterly cold
weather, little or no food, rampant disease (Tattersall 2006). McLardy survived and
eventually made it back to Britain.
Apart from plunging them into work that would help to shape their future careers, the
war had a strong formative affect on the general attitudes and aspiration of many
members. In a way that would just not happen in peace time, many were given huge
responsibilities and the freedom to follow their own initiative in solving their assigned
problems. (For a while, at barely 30 years of age, Pringle was in charge of all air-born
radar development in Britain – for his war service he was awarded a MBE and the
American Medal of Freedom with Bronze Palm.)
Craik
From the midst of this war-time interdisciplinary problem solving there emerged a
number of publications that were to have a galvanising affect on the development of
British cybernetics. These included K.J.W. Craik’s slim volume, The Nature of
Explanation, which first appeared in 1943 (Craik 1943). Bates’ hastily scrawled notes
for his introduction to the first meeting of the Ratio Club, a few lines on one side of a
scrap of paper, include a handful of phrases under the heading Membership. Of these
only one is underlined. In fact it is underlined three times. The phrase is ‘No Craik’.
Kenneth Craik was a Scottish psychologist of singular genius, who after many years
of neglect, is remembered now as a radical philosopher, a pioneer of the study of
human machine interfaces, a founder of cognitive psychology and as a father of
13
cybernetics thinking. His story is made particularly poignant by his tragic and sudden
demise at the age of 31 on the last day of the war in Europe. He was killed in a traffic
accident while cycling through Cambridge. He had recently been appointed the first
director of the Medical Research Council’s prestigious Applied Psychology Unit. He
was held in extremely high regard by Bates and the other Ratio members, so the ‘No
Craik’ was a lament.
After studying Philosophy at Edinburgh University, in 1936 he began a PhD. in
psychology and physiology at Cambridge. Here he came under the influence of
pioneering head of psychology Sir Frederick Bartlett. His love of mechanical devices,
and his skills as a designer of scientific apparatus, no doubt informed the radical
thesis of his classic 1943 book, published in the midst of his war work on factors
affecting the efficient operation and servicing of artillery machinery. Noting that ‘one
of the most fundamental properties of thought is its power of predicting events’
(Craik, ibid, p.50), Craik suggests that such predictive power is ‘not unique to minds’.
Indeed, although the ‘flexibility and versatility’ of human thought is unparalleled, he
saw no reason why, at least in principle, such essential properties as recognition and
memory could not be emulated by a man-made device. He went even further by
claiming that the human mind is a kind of machine that constructs small-scale models
of reality that it uses to anticipate events. In a move that anticipated Wiener’s
Cybernetics by five years, as well as foreshadowing the much later fields of Cognitive
Science and AI, he viewed the proper study of mind as an investigation of classes of
mechanisms capable of generating intelligent behaviour both in biological and nonbiological machines. Along with Turing, who is acknowledged in the introduction to
Wiener’s Cybernetics, and Ashby, who had begun publishing on formal theories of
adaptive behaviour in 1940 (Ashby 1940), Craik was a significant, and largely
forgotten, influence on American cybernetics. Both Wiener and McCulloch
acknowledged his ideas, quoting him in an approving way, and the later Artificial
Intelligence movement, founded by McCarthy and Minsky, was to a degree based on
the idea of using digital computers to explore Craik’s idea of intelligence involving
the construction of small-scale models of reality (see the original proposal for the
1956 Dartmouth Summer Project on AI for an explicit statement of this: McCarthy
1955). Many members of the Ratio Club, a high proportion of whom had connections
with Cambridge University, were influenced by Craik and held him in great esteem;
in particular Bates and Hick, who had both worked closely with him, Walter, who
cited war-time conversations with Craik as the original inspiration for the
development of his tortoises (Walter 1953, p.125), and Uttley, who Bates credited
with giving Craik many of his ideas (Bates 1945). Indeed, in a 1947 letter to Lord
Adrian, the charismatic Nobel prize winning head of physiology at Cambridge, Grey
Walter refers to the American Cybernetics movement as “… thinking on very much
the same lines as Kenneth Craik did, but with much less sparkle and humour.” (Walter
1947) Had he survived, there is no doubt Craik would have been a leading member of
the club. In fact John Westcott’s notes from the inaugural meeting of the club show
that there was a proposal to call it the Craik Club in his honour (Westcott 1949-53).
Existing relationships
Although the Ratio Club was the first regular gathering of this group of like-minded
individuals, certain members had interacted with each other for several years prior to
its founding, often in work or discussion with a distinct cybernetic flavour. For
14
instance, Bates and Hick had worked with Craik on war-time research related to
visual tracking in gunnery and the design of control systems in tanks. In the months
after Craik’s untimely death, they had been involved in an attempt to edit his notes for
a paper eventually published as 'Theory of the Human Operator in Control Systems'
(Craik 1948).
Ashby also was familiar with Craik’s ideas. In 1944 he wrote to Craik after reading
The Nature of Explanation (Craik 1943). As intimated earlier, the central thesis of
Craik’s book is that ‘thought models, or parallels, reality’ (Craik, ibid, p.57). Neural
mechanisms, somehow acting as ‘small-scale models’ of external reality, could be
used to “try out various alternatives, conclude which is the best of them, react to
future situations before they arise, utilise the knowledge of past event in dealing with
the present and future, and in every way to react in a much fuller, safer, and more
competent manner to the emergencies that face it.”(Craik, ibid, p.61). Today this is a
familiar idea, but Craik is widely acknowledged as the first thinker to articulate it in
detail. Ashby wrote to Craik to suggest that he needed to use terms more precise than
‘model’ and ‘paralleling’, putting forward group theory, in particular the concept of
isomorphism of groups, as a suitably exact language for discussing his theories
(Ashby 1944). He went on to state rather optimistically that “I believe ‘isomorphism’
is destined to play the same part in psychology that, say, velocity does in physics, in
the sense that one can’t get anywhere without it.” Craik took this suggestion seriously
enough to respond with a three page letter on the nature of knowledge and
mathematical description which resulted in a further exchange of letters revealing a
fair amount of common ground in the two men’s views on what kind of knowledge
science could communicate. Craik was ‘much interested to hear further’ (Craik 1944)
of Ashby’s theories alluded to in the following paragraph in which he introduces
himself:
“Professionally I am a psychiatrist, but am much interested in mathematics,
physics and the nervous system. For some years I have been working on the
idea expressed so clearly on p.115: “It is possible that a brain consisting of
randomly connected impressionable synapses would assume the required
degree of orderliness as a result of experience …” After some years’
investigation of this idea I eventually established that this is certainly so,
provided that by “orderly” we understand “organised as a dynamic system so
that the behaviour produced is self-preservative rather than self-destructive”.
The basic principle is quite simple but the statement in full mathematical
rigour, which I have recently achieved, tends unfortunately to obscure this
somewhat.” (Ashby 1944)
In Ashby’s talk of self-preservative dynamic systems we can clearly recognise the
core idea he would continue to develop over the next few years and publish in Design
for a Brain (Ashby 1952a). In that book he constructed a general theory of adaptive
systems as dynamical systems in which ‘essential’ variables (such as heart rate and
body temperature in animals) must be kept within certain bounds in the face of
external and internal changes or disturbances. This work, which preoccupied Ashby
for the early years of the Ratio Club, is discussed later in more detail in Section 5.
Ashby corresponded with several future members of the club in the mid 1940s. For
instance, in 1946 William Hick wrote to Ashby after reading his note on equilibrium
15
systems in the American Journal of Psychology (Ashby 1946). Hick explained that he
too was “… trying to develop the principles of ‘Analytical Machines’ as applied to
the nervous system…” (Hick 1947a) and requested copies of all Ashby’s papers. The
pair corresponded over the mathematical details of Ashby’s theories of adaptation
with Hick declaring himself “… not entirely happy with your conclusion that a
sequence of breaks, if it continues long enough, will eventually, by chance, lead to a
stable equilibrium configuration.” (Hick 1947b). Hick was referring to an early
description of what would later appear in Design for a Brain as postulated step
mechanisms that would, following a disturbance that pushed any of the system’s
essential variables out of range, change the internal dynamics of an adaptive machine
until a new equilibrium was established (i.e. all essential variables were back in range
– see Section 5 for further details). Ashby agrees, explaining that he had no rigorous
proof but “… has little doubt of its truth in a rough and ready , practical way.”
(Ashby 1947). A year later the Homeostat machine would provide an existence proof
that these mechanisms could work. But Hick had homed in on an interesting and
contentious aspect of Ashby’s theory. By the time Design for a Brain was published,
Ashby talked about step mechanisms in very general terms, stating that they could be
random but did not ascribing absolute rigid properties to them, hence leaving the door
open for further refinements. This correspondence foreshadows the kind of probing
discussions that were to form the central activity of the Ratio Club, debates that
sometimes spilled out onto the pages of journals (see Section 5 for an example of
this).
During the war there had been considerable interaction between researchers at the
various military sites and several members had originally met through that route. For
instance, one of the things Uttley worked on at TRE was computer-aided target
tracking, as well as building the first British airborne electronic navigation computer
(Uttley 1982). This work on early computing devices brought him into contact with
both Gold and Turing.
Several members had been friends or acquaintances at Cambridge (e.g. Pringle and
Turing were contemporaries, as were Barlow and Merton who had both been tutored
by Rushton). Others met at workshops and conferences in the years leading up to the
founding of the club. Those involved in EEG work (Walter, Bates, Dawson, Shipton)
were all well known to each other professionally (Walter and Dawson had together
laid the foundations for clinical uses of EEG; a paper they wrote together in 1944
(Dawson and Walter 1944) was still used in the training of EEG practitioners in the
1980s). Ashby had interacted with Walter for some time, not least because their
research institutes were nearby. So, by the time the Ratio Club started, most members
had at least passing familiarity with some, but by no means all, of the other’s ideas.
3.3 The way forward
For two or three years prior to the founding of the club there had been a gradual
increase in activity, on both sides of the Atlantic, on new approaches to machine
intelligence, as well as renewed interest in associated mechanistic views of natural
intelligence. In Britain much of that activity involved future Ratio members. The
phrase Bates used in his initial letters of invitation to the founders of the club, that he
wished to bring together people who ‘… had Wiener’s ideas before Wiener’s book
16
appeared …’, may have been slightly gung-ho, but in a draft for an article for the
British Medical Journal in 1952, Bates explained himself a little more:
"Those who have been influenced by these ideas so far, would not
acknowledge any particular indebtedness to Wiener, for although he was the
first to collect them together under one cover, they had been common
knowledge to many workers in biology who had contacts with various types of
engineering during the war." (Bates 1952b)
It is likely that Bates was mainly thinking of chapters III, IV and V of Cybernetics,
those on ‘time series, information and communication’, ‘feedback and oscillation’ and
‘computing machines and the nervous system’. Certainly many biologist had become
familiar with feedback and its mathematical treatment during the war, and some had
worked on time series analysis and communication in relation to radar (some of their
more mathematical colleagues would have been using some of Wiener’s techniques
and methods that were circulating in technical reports and draft papers – quite literally
having Wiener’s ideas before the book appeared). Most felt that the independent
British line of research on computing machines and their relationship to the nervous
system was at least as strong as the work going on in the USA (important strands of
which were based on prior British work such as that of Turing) (Barlow 2001).
Indeed, many were of the opinion that the central hypothesis of cybernetics was that
the nervous system should be viewed as a self-correcting device chiefly relying on
negative-feedback mechanisms (Wisdom 1951). This concept had first been
introduced by Ashby in 1940 (Ashby 1940) and then independently by Wiener and
colleagues three years later (Rosenblueth et al. 1943). The development of this idea
was the central, all-consuming focus of Ashby’s work until the completion of Design
for a Brain, which set out his theories up to that point. It is interesting that Ashby’s
review of Cybernetics (Ashby 1949b) is quite critical of the way in which the core
ideas of the book are presented.
Perhaps the following passage from the introduction to Cybernetics pricked Bates’
sense of national pride and acted as a further spur:
“In the spring of 1947 … [I] spent a total of three weeks in England, chiefly as
a guest of my old friend J.B.S. Haldane. I had an excellent chance to meet
most of those doing work on ultra-rapid computing machines … and above all
to talk over the fundamental ideas of cybernetics with Mr. Turing … I found
the interest in cybernetics about as great and well informed in England as in
the United States, and the engineering work excellent, though of course
limited by the smaller funds available. I found much interest and
understanding of its possibility in many quarters … I did not find, however,
that as much progress had been made in unifying the subject and in pulling the
various threads of research together as we had made at home in the States.”
(Wiener 1948, p.23)
However, whatever the views on Wiener’s influence – and the more mathematical
members will surely have recognised his significant technical contributions – it is
clear that all those associated with the Ratio Club agreed that Shannon’s newly
published formulation of information theory, partly built on foundations laid by
17
Wiener, was very exciting and important. The time was ripe for a regular gathering to
develop these ideas further.
4. Club meetings
The London district of Bloomsbury often conjures up images of free thinking
intellectuals, dissolute artists and neurotic writers – early twentieth century bohemians
who, as Dorothy Parker once said, ‘lived in squares and loved in triangles’. But it is
also the birth place of neurology, for it was here, in 1860, that the first hospital in the
world dedicated to the study and treatment of diseases of the nervous system was
established. By the late 1940s The National Hospital for Nervous Diseases was
globally influential and had expanded to take up most of one side of Queen’s Square.
It was about to become regular host to the newly formed group of brilliant and
unconventional thinkers.
In 1949 London witnessed the hottest September on record up to that point, with
temperatures well above 90F. In fact the entire summer had been a mixture of
scorching sunshine and wild thunderstorms. So it was an unseasonably balmy evening
on the 14th of that month when a gang of scientists, from Cambridge in the east and
Bristol in the west, descended on the grimy bombed-out capital, a city slowly
recovering from a war that had financially crippled Britain. They converged on the
leafy Queen’s Square and assembled in a basement room of the hospital at 6.30 in the
evening. After sherries, the meeting started at 7:00. Bates’ notes for his introduction
to this inaugural gathering of the club show that he spoke about how the club
membership was drawn from a network based around his friends, and so was
somewhat arbitrary, but that there had been an attempt to strike a balance between
biologists and non-biologists (Bates 1949e). He then went on to make it clear that the
club was for people who were actively using cybernetic ideas in their work. At that
point there were 17 members but he felt there was room for a few more. (The initial
membership comprised: Ashby, Barlow, Bates, Dawson, Gold, Hick, Little, Mackay,
McLardy, Merton, Pringle, Shipton, Sholl, Slater, Uttley, Walter, Westcott). He
pointed out that there were no sociologists, no northerners (for example from
Manchester University) and no professors. Possible names for the club were discussed
(see Section 3.1) before Bates sketched out how he thought meetings should be
conducted. In this matter he stressed the informality of the club – that members
should not try and impose ‘direction’ or employ ‘personal weight’. All agreed with
this sentiment and endorsed his ‘no professors’ rule – scientists who were regarded to
be senior enough to inhibit free discussion were not eligible for membership.
Warren McCulloch then gave his presentation, Finality and Form in Nervous Activity,
a popular talk that he had first given in 1946 – perhaps not the best choice for such a
demanding audience. Correspondence between members reveals almost unanimous
disappointment in the talk. Bates set out his own reaction to its content (and style) in a
letter to Grey Walter:
"Dear Grey,
18
Many thanks for your letter. I had led myself to expect too much of
McCulloch and I was a little disappointed; partly for the reason that I find all
Americans less clever than they appear to think themselves; partly because I
discovered by hearing him talk on 6 occasions and by drinking with him in
private on several more, that he had chunks of his purple stuff stored parrotwise. By and large however, I found him good value." (Bates 1949f)
Walter replied to Bates apologizing for not being present at the meeting (he was the
only founding member unable to attend). This was due to the birth of a son, or as he
put it “owing to the delivery of a male homeostat which I was anxious to get into
commission as soon as possible.” (Walter 1949) He went on to tell Bates that he has
had ‘an amusing time’ with McCulloch who had travelled on to Bristol to visit him at
the Burden Institute. In reference to Bates’ view on McCulloch’s talk, he comments
“… his reasoning has reached a plateau … flowers that bloom on this alp are worth
gathering but one should keep one’s eyes on the heights.”
A buffet dinner with beer followed the talk and then there was an extended discussion
session. The whole meeting lasted about three hours. Before the gathering broke up,
with some rushing off to catch last trains out of London and others joining McCulloch
in search of a night cap, John Pringle proposed an additional member. Echoing the
suggestion made in his written reply to Bates’ original invitation to join the club,
Pringle put forward the idea that a mathematician or two should be invited to join to
give a different perspective and to ‘keep the biologists in order’. He and Gold
proposed Alan Turing, a suggestion that was unanimously supported. Turing gladly
accepted and shortly afterwards was joined by fellow mathematician Philip
Woodward, who worked with Uttley. At the same time leading Cambridge
neurobiologist William Rushton, who was well know to many members, was added to
the list. The following passage from a circular Bates sent to all members shortly after
the first meeting shows that the format for the next few sessions had also been
discussed and agreed:
“It seems to be accepted that the next few meetings shall be given over to a
few personal introductory comments from each member in turn. Assuming we
can allow two and a half hours per meeting, eighteen members can occupy an
average of not more than 25 minutes each. The contributions should thus
clearly be in the nature of an aperitif or an hors d’oeuvres – the fish, meat and
sweet to follow at later meetings.” (Bates 1949g)
Regardless of reactions to the opening talk, there was great enthusiasm for the
venture. The club was well and truly born.
Following the inaugural meeting in September 1949, the club convened regularly until
the end of 1954. There was a further two day meeting and a single evening session in
1955 and a final gathering in 1958 after the now classic Mechanization of Thought
Processes symposium organized by Uttley at the National Physical Laboratory in
Teddington (Blake and Uttley 1959).
Table 1 shows the full list of known Ratio Club meetings. This has been compiled
from a combination of individual meeting notices found in the Bates archive,
surviving members’ personal records and a list of meetings made by Bates in the mid
19
1980s. There are inconsistencies between these sources, but through cross-referencing
with notes made at meetings and correspondence between members this list is
believed to be accurate. It is possible that it is incomplete – if so, only a very small
number of additional meetings could have occurred.
Figure 1: The main entrance to the National Hospital, Queen’s Square in 2002. Ratio Club
meetings were held in a room in the basement.
20
Meeting
1
2
3
4
5
Date
14 Sep. 1949
18 Oct. 1949
17 Nov. 1949
15 Dec. 1949
19 Jan. 1950
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
16 Feb. 1950
16 Mar. 1950
21 Apr. 1950
18 May 1950
22 June 1950
18 July 1950
21 Sep. 1950
2 Oct. 1950
7 Dec. 1950
22 Feb. 1951
5 Apr. 1951
31 May 1951
26 July 1951
1 Nov. 1951
21 Dec. 1951
21
8 Feb. 1952
22
20 Mar. 1952
23
24
25
26
27
2-3 May 1952
19 June 1952
31 July 1952
24-25 Oct. 1952
6 Nov. 1952
28
11 Dec. 1952
29
30
19 Feb. 1953
7 May 1953
31
2 July 1953
32
22 Oct. 1953
33
11 Feb. 1954
34
17 June 1954
35
25 Nov. 1954
36
37
6-7 May 1955
15th Sep. 1955
38
27 Nov. 1958
Speakers, title
Warren McCulloch, Finality and Form in Nervous Activity
Introductory talks from Sholl, Dawson, Mackay, Uttley.
Introductory talks from Gold, Bates, McLardy
Introductory talks from Pringle, Merton, Little, Hick; Grey Walter
E. Slater, Paradoxes are Hogwash; D. Mackay, Why is the Visual
World Stable?
Introductory talks from Shipton, Slater, Woodward
Introductory talks from Ashby, Barlow
Introductory talks from Wescott, Turing
Pattern Recognition, Walter, Uttley, Mackay, Barlow, Gold
Elementary basis of Information Theory, Woodward
Concept of Probability, Gold, Mackay, Sholl
Noise in the Nervous System, Pringle
Meeting at London Symposium on Information Theory
Educating a Digital Computer, Turing
Adaptive Behaviour, Walter
Shape and Size of Nerve Fibres, Rushton
Statistical Machinery, Ashby
Telepathy, Bates
On Popper: What is happening to the universe?, Gold
Future Policy; discussion on The possibility of a scientific basis of
ethics opened by Slater; discussion on A quantitative approach to
brain cell counts opened by Sholl.
The Chemical Origin of Biological Form, Turing; The Theory of
Observation, Woodward
Pattern Recognition, Uttley; Meaning in Information. Theory,
Mackay
Special Meeting at Cambridge (organised by Pringle).
Memory, Bates; The Logic of Discrimination, Westcott
The Size of Eyes, Barlow; American Interests in Brain Structure, Sholl
Special Bristol Meeting at Burden Inst. (cancelled)
Design of Randomizing Devices, Hick; On Ashby’s Design for a
Brain, Walter
Perils of Self-awareness in Machines, Mackay; Sorting Afferent from
Efferent Messages in Nerves, Merton
Pattern Discrimination in the Visual Cortex, Uttley and Sholl
Absorption of Radio Frequencies by Ionic Materials, Little; The
Signal to Noise Problem, Dawson
Warren McCulloch: Discussion of topics raised in longer lectures
given by McCulloch at UCL in previous week
Demonstration and Discussion of the Toposcope, Shipton; Principles
of Rational Judgement, Good
Discussion on How does the nervous system carry information?;
Guest talk: Observations on Hearing Mechanisms, Whitfield and
Allanson
Servo control of Muscular Movements, Merton; Introduction to Group
Theory, Woodward,
Negative Information, Slater and Woodward; Guest talk: Development
as a Cybernetic Process, Waddington
Special meeting West Country (TRE, Barnwood, Burden)
Discussion meeting after 3rd London Symposium on Information
Theory. Many guests from USA.
Final reunion meeting after NPL Mechanisation of Thought Processes
Symposium
Table 1: Known Ratio Club Meetings
21
The order of member’s introductory talks was assigned by Bates using a table of
random numbers. Due to overruns and some people being unable to attend certain
meetings, the actual order in which they were given may have been slightly different
from that shown in the table. However, they did occupy the dates indicated.
The format of the opening meeting – drinks, session, buffet and beer, session, coffee –
seems to have been adopted for subsequent meetings. Member’s introductory talks,
which highlighted their expertise and interests, typically focused on some aspect of
their current research. John Westcott’s notebook reveals that a wide range of topics
were discussed (Westcott 1949-53): Scholl talked about the need to construct an
appropriate mathematics to shed light on the physiology of the nervous systems;
Dawson described on-going work on eliminating noise from EEG readings; Mackay
argued for a more complex description of information, both philosophically and
mathematically, claiming that it cannot be adequately defined as a single number;
Uttley sketched out the design for a digital computer he was working on at TRE; Gold
illustrated his more general interest in the role of servomechanism in physiology by
describing his work on a radical new theory of the functioning of the ear, which
postulated a central role for feedback (Gold (2002) later remembered that at the time
the Ratio Club was the only group that understood his theory); Bates talked about
various levels of description of the nervous system; McLardy described recent
research in invasive surgical procedures in psychiatry; Merton outlined his work on
using cybernetic ideas to gain a better understanding of how muscles work; Walter
described his newly constructed robotic tortoises, sketching out the aims of the
research and early results obtained (see Section 5 for further discussion of this work);
Woodward talked about information in noisy environments; Little discussed the
scientific method and the difficulty of recognizing a perfect theory; Hick outlined his
research on reaction times in the face of multiple choices – the foundations of what
would later become known as Hick’s law (Hick 1952) which makes use of
information theory to describe the time taken to make a decision as a function of the
number of alternatives available (see Section 2 for a brief statement of the law);
Ashby talked about his theories of adaptive behaviour and how they were illustrated
by his just-finished Homeostat device (see Section 5 for further discussion of this
work); Barlow outlined the research on the role of eye movement in generating visual
responses that he was conducting at this early stage of his career (see the interview
with Barlow later in this volume for further details of this work), and Westcott talked
a little about his background in radar and his work with Wiener at MIT before
outlining his mathematical work on analysing servomechanisms, emphasising the
importance of Wiener’s theory of feedback systems on which he was building. After
each of the presentations discussion from the floor took over.
After the series of introductory talks, the format of meetings changed to focus on a
single topic, sometimes introduced by one person, sometimes by several. Prior to this,
Ashby circulated two lists of suggested topics for discussion; an initial one on 18th
February 1950 (Ashby 1950a) and a refined version dated 15th May 1950 (Ashby
1950b). They make fascinating reading; giving an insight into Ashby’s
preoccupations at the time. The refined list is reproduced here. Many of the questions
are still highly pertinent today.
22
1. What is known of 'machines' that are defined only statistically? To what extent is
this knowledge applicable to the brain?
2. What evidence is there that 'noise' (a) does, (b) does not, play a part in brain
function?
3. To what extent can the abnormalities of brains and machines be reduced to
common terms?
4. The brain shows some indifference to the exact localisation of some of its
processes: to what extent can this indifference be paralleled in physical systems? Can
any general principle be deduced from them, suitable for application to the brain?
5. From what is known about present-day mechanical memories can any principle be
deduced to which the brain must be subject?
6. To what extent do the sense-organs' known properties illustrate the principles of
information-theory?
7. Consider the various well known optical illusions: what can information-theory
deduce from them?
8. What are the general effects, in machines and brains of delay in the transmission of
information?
9. Can the members agree on definitions, applicable equally to all systems biological, physiological, physical, sociological - cf: feedback, stability, servomechanism.
10. The physiologist observing the brain and the physicist observing an atomic system
are each observing a system only partly accessible to observation: to what extent can
they use common principles?
11. The two observers of 10, above, are also alike in that each can observe his system
only by interfering with it: to what extent can they use common principles?
12. Is 'mind' a physical 'unobservable'? If so, what corollaries may be drawn?
13. What are the applications, to cerebral processes, of the thermodynamics of open
systems?
14. To what extent can the phenomena of life be imitated by present-day machines?
15. To what extent have mechanisms been successful in imitating the conditioned
reflex? What features of the C.R. have conspicuously not yet been imitated?
16. What principles must govern the design of a machine which, like the brain, has to
work out its own formulae for prediction?
17. What cerebral processes are recognisably (a) analogical, (b) digital, in nature?
18. What conditions are necessary and sufficient that a machine built of many
integrated parts should be able, like the brain, to perform an action either quickly or
slowly without becoming uncoordinated?
19. Steady states in economic systems.
20. What general methods are available for making systems stable, and what are their
applications to physiology?
21. To what extent can information-theory be applied to communication in insect and
similar communities?
22. To what extent are the principles of discontinuous servo-mechanisms applicable to
the brain?
23. What re-organisation of the Civil Service would improve it cybernetically?
24. What economic 'vicious circles' can be explained cybernetically?
25. What re-organisation of the present economic system would improve it
cybernetically?
26. To what extent can information-theory be applied to the control exerted
genetically by one generation over the next?
23
27. Can the members agree on a conclusion about extra-sensory perception?
28. What would be the properties of a machine whose 'time -was not a real but a
complex variable? Has such a system any application to certain obscure, i.e.
spiritualistic, properties of the brain?
The last topic on the initial list is missing from the more detailed second list: ‘If all
else fails: The effect of alcohol on control and communication, with practical
work.’ This suggestion was certainly taken up, as it appears were several others:
shortly after the lists appeared Pringle gave a talk on the topic of suggestion 2
(meeting 12), as did Walter on 14 and 15 (meeting 15). Topic 27 came up in talks by
Bates and Good (meetings 18 and 32). Issues relating to many of the other
suggestions often arose in group discussions, being in areas of great interest to many
members (e.g. topics 6-13, 16-18 and 26). In particular, Barlow recalls much
discussion of topic 17 (Barlow 2007).While Ashby’s publications and notebooks
make it clear that some of the suggestions are based on the central research questions
he was grappling with at the time (e.g. 1, 18, 20, 22), it is very likely that some of the
others arose from issues brought up by other members in their introductory talks. In
the mid 1980s Bates made some notes for a planned article on the Ratio Club (Bates
1985), a plan which unfortunately did not come to fruition. However, among these
scant jottings is mention of Ashby’s lists, which further suggests that they did play a
role in shaping the scope of topics discussed.
Members often volunteered to give talks, but when he felt it was necessary Bates
actively controlled the balance of topics by persuading particular members to give
presentations. Sometimes there were requests from members for particular subjects to
be discussed or particular people to give talks on certain topics. Looking through the
list of subjects discussed, many are still extremely interesting today; at the time they
must have been positively mouth watering.
At the end of 1950, after meeting him at the first London Symposium on Information
Theory, Bates invited I.J. ‘Jack’ Good along to the next meeting as his guest. The
speaker was Turing, Good’s friend and war-time colleague. This was a particularly
lively meeting and after it Good wrote to Bates expressing how much he had enjoyed
the evening and apologizing for being too vociferous. He wondered “would there be
any serious objection to my becoming a member?” (Good 1950a). Bates replied that
“the club has been going for a year, and is entirely without any formal procedures.
New members join by invitation, but I think personally you would be a great asset,
and hope you will be able to come as my guest to some future meetings, so that
perhaps my view will become consensus!”(Bates 1950a). Bates’ view obviously did
hold sway as Good became the twenty-first member of the Club. Perhaps it was
thought a third mathematician was needed to help the other two keep the biologists in
order. Partly because of the size of the room used for meetings, and partly because
Bates had firm ideas on the kind of atmosphere he wanted to create and who were the
‘right sorts’ to maintain it, the membership list remained closed from that point.
For the first year meetings were monthly and were all held at the National Hospital in
Queen’s Square. From mid 1950 until the end of 1951 the frequency of meetings
dropped slightly and in the second half of 1951 attendance started to fall. This was
mainly due to the not inconsiderable time and expense incurred by members who
were based outside London every time they came to a meeting. In October 1951
24
Woodward had written to Bates explaining that he had to take part of his annual leave
to attend meetings (Woodward 1951); the following month Walter wrote to explain
that he had difficulty in covering the expenses of the trips to London necessary for
Ratio gatherings. He suggested holding some meetings outside London in member’s
labs, pointing out that this would also allow practical demonstrations as background
for discussion (Walter 1951).
Indeed the return journey to Bristol could be quite a hike. Janet Shipton remembers
waiting up to greet her husband, Harold, on his return from Ratio meetings, “He
would get back in the dead of night, the smell of train smoke on his clothes.” (Shipton
2002)
At the December 1951 meeting of the club Bates called a special session to discuss
future policy. Beforehand he circulated a document in which he put down his thoughts
on the state of the club. Headed ‘The Ratio Club’, the document opened by stating
that “looked at in one way, the Club is thriving - in another way it is not. It is thriving
as judged by the suggestions for future activities.” (Bates 1951). These suggestions
are listed as requests for specific talks by Woodward (on the theory of observation)
and Hick (on the rate of gain of information), an offer of a talk on morphogenesis by
Turing, as well as various suggestions for discussion topics (all of these suggestions,
offers and requests were taken up in subsequent meetings). Bates goes on: “In
addition to this, we have in pigeon-holes a long list sent in by Ashby of suitable
topics; various suggestions for outside speakers; and a further suggestion that
members should collaborate in writing different chapters to a book on the lines of
‘Cybernetics’, but somewhat tidier.” Sadly, this intriguing book idea never came to
fruition. He then explains the cause for concern:
“Looked at in another way, the Club is ailing. For the past three meetings, half
or more of the members have been absent. This half have been mostly those
who live out of London - the most reasonable inference clearly is that a single
evening's meeting does not promise to be a sufficient reward for the
inconvenience and expense of getting to it. In addition one member has
pointed out that if expenses cannot be claimed the night's absence is counted
against the period of his annual leave! The whole point of the Club is to
facilitate contacts between people who may have something to contribute to
each other, and who might not otherwise come together, and it would seem
that some change in its habits may be indicated.” (Bates 1951)
Bates then list some suggested courses of action for discussion at the next meeting.
These ranged from having far fewer, but longer, meetings to doubling the
membership.
It was decided that there would be six or seven meeting a year, four or five in London
and two elsewhere. The meetings would start earlier to allow two papers. A novel
suggestion by Philip Woodward was also taken up: to start a postal portfolio – a
circulating package of ideas – ‘to be totally informal and colloquial’. Bates prepared a
randomized order of members to dictate the route the portfolio would travel.
This new regime was followed from the first meeting of 1952 until the club
disbanded, and seemed to go a good way towards solving the problems that prompted
25
its instigation. The typical meeting pattern was now to gather at 4:30 for tea followed
by the first talk and discussion, then a meal and drinks followed by the second talk
and discussion.
Most Ratio talks were based on current research and were often early outings for
highly significant work, sometimes opening up new areas of enquiry that are still
active today. For instance, Turing’s talk on Educating a Digital Computer, in
December 1950, was on the topics covered by his seminal Mind paper of that year
(Turing 1950) which introduced The Turing Test and is regarded as one of the key
foundational works of machine intelligence. As the title suggests, that talk focused on
how an intelligent machine might be developed – Turing advocated using adaptive
machines that might learn over their lifetime and also over generations by employing
a form of artificial evolution. This meeting is remembered as being particularly good
with Turing on top form stimulating a scintillating extended discussion (Bates 1950b).
Turing’s 1952 talk on biological form was another gem, describing his as yet
unpublished work on reaction-diffusion models of morphogenesis (Turing 1952)
which showed how pattern and form could emerge from reaction-diffusion systems if
they are appropriately parameterized (a role he hypothesized might be taken on by
genes). Apart from launching new directions in theoretical biology, this work was
pioneering in its use of computer modeling and was to prove extremely influential.
There is not enough space to describe all the important work discussed at club
meetings but further summaries will be scattered at appropriate places throughout the
rest of this paper.
As well as research talks, there were also various ‘educational’ presentations, usually
requested by the biologists. For instance, Woodward gave several on information
theory, which gave the biologists very early access to important new ways of
thinking. By all accounts Woodward was an extremely good lecturer, blessed with a
gift for insightful exposition (this is evident in his 1953 book Probability and
Information Theory, with Applications to Radar – Woodward 1953, still regarded by
some theorists as one of the most profound works in the area since Shannon’s original
papers.) Barlow was particularly influenced by these exciting new ideas and became a
pioneer of the use of information theory as a theoretical framework to understand the
operation of neural systems, particularly those associated with vision. This theoretical
framework either directly or indirectly underpinned many of Barlow’s very important
contributions to neuroscience. He regards the Ratio Club as one of the most important
formative influences on his work and sees “much of what I have done since as
flowing from those evening meetings” (Barlow 2001) (see the interview with Barlow
later in this volume for further discussion of this point). In a similar spirit there were
lectures on probability theory from Gold and Mackay and on the emerging field of
control theory from Westcott.
In 1952 two extended out-of-London meetings were planned, one in Cambridge in
May and one in Bristol in October. The Cambridge meeting was organized by Pringle
and was held from Friday afternoon to Saturday morning in his college, Peterhouse.
After drinks and dinner Pringle led a session on Processes Involved in the Origin of
Life. Correspondence after the meeting mentions that this session was captured on a
tape recorder, although the recording has not yet come to light. The next day visits
were arranged to various labs including: Cavendish (physics), led by Gold; Zoology;
Physiology, led by Rushton; Psychology, led by Hick; and Mathematics. The
26
photograph shown in Figure 2 was taken at this meeting, quite possibly after the predinner sherries mentioned on the invitation sent out to Club members. Not everyone
was able to attend and several of those in the photograph are guests. A limited number
of guests were allowed at most meetings and over the years various distinguished
visitors took part in club gatherings. As well as McCulloch, Pitts and Shannon, these
included J.Z. Young, the leading anatomist and neurologist, who attended several
meetings, C.H. Waddington, the pioneering theoretical biologist and geneticist, and
Giles Brindley who became a distinguished neuroscientist and was David Marr’s PhD
supervisor. Jack Good once brought along the director of the NSA, home to the
USA’s code breakers and makers, who he knew through his work for British
Intelligence. That particular meeting was on probability and included prolonged
discussions of experiments claiming to give evidence for ESP. Following the 1955
London Symposium on Information Theory, a special club meeting involved a host of
leading lights from the world of information theory and cybernetics, many from
overseas. These included: Peter Elias, JCR Licklider, Warren McCulloch, Oliver
Selfridge, Benoît Mandelbrot and Colin Cherry. Records are sketchy on this matter,
but it is likely that many other luminaries of the day took part in other meetings.
Figure 2: Some members of The Ratio Club with guests outside Peterhouse,
University of Cambridge, May 1952. The photograph was organized by Donald
Mackay. Back row (partly obscured): H. Shipton, J. Bates, W. Hicks, J. Pringle,
D. Scholl, J. Westcott, D. Mackay. Middle row: G. Brindley, T. McLardy. W.R.
Ashby, T. Gold, A. Uttley. Front row: A. Turing, G. Sutton, W. Rushton, G.
Dawson, H. Barlow. Courtesy Wellcome Library for the History and
Understanding of Medicine.
27
The Bristol meeting was to be held at the Burden Neurological Institute starting at
noon on Friday 24th October 1952 and running into the next day but it seems to have
been cancelled at the last minute dues to heavy teaching commitments preventing a
substantial number of members from attending. The talks and demonstrations planned
for this meeting were moved into later club meetings. These included Grey Walter
opening a discussion on Mechanisms for Adaptive Behaviour which focused on
simulation of learning by man-made devices, and in particular on the issues raised in
Ashby’s recently published book Design for a Brain, and a presentation by Shipton
on the Toposcope, the world’s first multi-channel EEG recording device. The
machine, developed by Shipton and Walter, was capable of building and displaying
bidimensional maps of the EEG activity over the brain surface and included frequency
and phase information.
From mid 1953 meeting became less frequent, with only three in 1954 and two in
1955. In 1955 the extended West-Country event finally happened, starting at TRE
Malvern on May 6th and then going the next day to the Burden Institute in Bristol via
Ashby’s Barnwood House lab. The meeting was primarily devoted to demonstrations
and discussions of work in progress at these locations. At TRE various devices from
Uttley’s group were on show. These included a ‘tracking simulator’, a novel apparatus
designed to provide a versatile means of setting up and studying problems relating to
a human operator working in a closed loop system. The device used a two gun
cathode ray tube and required the operator to track a moving dot by controlling a
second dot with a joystick. Also on show were Uttley’s systems for automatically
classifying spatial and temporal patterns and pioneering electronic and hydraulic
systems capable of inference using principles from conditional probability.
To reach the next leg of the multi-site meeting, Philip Woodward recalls travelling
across country in a Rolls Royce Barlow had borrowed from his brother. As they
hurtled towards ‘Ashby’s lunatic asylum’, Rushton diagnosed the exact form of
Woodward’s colour blindness by getting him to describe the spring flowers he could
see on the verges (Woodward 2002).
At Barnwood House, Ashby demonstrated his Dispersive and Multi-stable System
(DAMS), and the Homeostat was available for those who were not already familiar
with it. As mentioned earlier, the Homeostat demonstrated the theories of adaptation
developed in Design for a Brain, where it is described in some detail. Although
Ashby had talked at earlier club meetings about the DAMS machine, this would have
been the first time that most members had seen it first hand. The DAMS device,
which is much less well know than the Homeostat – mainly because Ashby was not
able to develop it sufficiently to fully demonstrate his theories – was intended to
explore possible learning behaviours of randomly connected non-linear components.
The motivation for this was the intriguing possibility that parts of the brain,
particularly the cortex, might be (at least partially) randomly wired. Ashby had been
developing the machine for some years and demonstrated the current version which
by then illustrated some interesting properties of ‘statistical machinery’. The
theoretical line started in this work resurfaced many years later in Gardner and
Ashby’s computational study of the stability of large interconnected systems (Gardner
and Ashby 1970). There is a nice anecdote about the machine which originates from
this 1955 meeting. Philip Woodward remembers being told, possibly apocryphally,
28
that when Ashby asked a local engineering firm to construct part of the device,
specifying random connections, they were so bemused, particularly since the order
was coming from Barnwood House Psychiatric Hospital, that they rang up to check
that Dr. Ashby was not in fact a patient (Woodward 2002).
The club was full of lively and strong personalities. Mackay, Turing and Walter were,
in their very different ways, brilliant speakers who each made scientific broadcasts for
BBC radio. Grey Walter, in particular, was something of a media personality with
appearances on popular radio quiz shows and early television programmes. He was a
larger than life character who liked to cultivate a certain image – that of a
swashbuckling man of the world. He was, as Harold Shipton noted in 2002, ‘a bugger
for the women’ (Shipton 2002). This reputation did him no favours with many in the
scientific establishment. Walter stood in marked contrast to Mackay, a fiery lay
preacher who had been brought up attending the Evangelical Free Church of Scotland
(one of the radical breakaway ‘wee free’ churches). Many who knew him have
remarked on a certain tension between his often radical scientific ideas about the
nature of intelligence and his strait-laced religiosity. Horace Barlow, a great friend of
Mackay and an admirer of his ideas, has noted that “his conviction that he had a
special direct line to a Higher Place, somehow slightly marred his work and prevented
him from becoming as well regarded as he should have been.” (Barlow 2002)
According to Barlow’s biographical memoir of Rushton (Barlow 1986), his former
tutor “cut a striking and influential figure .. was argumentative, and often an
enormously successful showman .. he valued the human intellect and its skilful use
above everything else.” Giles Brindley, a guest at several meetings, remembers that
Barlow and Gold were very active in discussions and that when occasionally a debate
got out of hand, Pringle would gently refocus the conversation (Brindley 2002).
Members came from a rich mix of social and educational backgrounds ranging from
privileged upbringings to the humblest of origins. Harold Shipton’s story is
particularly remarkable. In the years before World War II he was plucked from the
life of an impoverished farm labourer by RAF talent scouts who were looking for
bright young men to train as radar operators. During training it quickly became
apparent that he had a natural gift for electronics which was duly exploited. After the
war, before he had been demobbed, he was sent to the Burden Institute to find out
what Grey Walter was doing with the suspiciously large amounts of surplus military
electronic equipment he was buying. He and Walter immediately hit it off and he
stayed. At the institute he met his future wife – Clement Attlee’s daughter Janet.
(Attlee was leader of the Labour party, Churchill’s deputy during the war and Prime
Minister of Britain from 1945-51.) Hence, at the west-country meeting, members of
the all-male Ratio Club were served tea by the Labour Prime Minister’s daughter.
Bates had created a powerful mix of individuals and ideas with just the right degree of
volatility. The result was that meetings were extremely stimulating and greatly
enjoyed by all. All the surviving members interviewed recalled the club with great
enthusiasm (Gold (2002) described meetings as “always interesting, often exciting.”).
Even those, such as Woodward and Westcott, who felt that they were net givers, in
terms of the direct intellectual influence of the club on members’ work, found
meetings a pleasure and were annoyed when they had to miss one.
29
The social atmosphere of the club was sometimes carried on in after-meeting parties.
Philip Woodward remembers that on one occasion some of the group reconvened on
the enormous Dutch sailing barge Pat Merton kept in St. Catherine’s docks on the
Thames. Merton had arranged for a pianist and oboeist to play on deck. St.
Catherine’s was a working dock in those days with a large sugar refinery that belched
out pungent fumes. As the night wore on and the drink flowed, the sugar strewn route
up the dock side steps to the toilet became more and more treacherous. (Woodward
2002)
The list in Table 1 shows that a number of external speakers were invited to give
presentations. Despite reactions to his talk in 1949, Warren McCulloch was asked
back in 1953 to open a discussion on his work, and attended other meetings as a guest
– members grew to appreciate his style.
As the club developed Ashby was keen to see it transform into a formal scientific
society – 'The Biophysical Society' or 'The Cybernetics Society' – with a more open
membership. His proposals for this were resisted. It seems that, for many members,
the informal atmosphere of the club, exactly as Bates had conceived it, was the most
important factor. When Ashby proposed that Professor J.Z. Young should be admitted
as a member, Sholl wrote to Bates in protest:
"I consider membership of the Club not only as one of my more pleasant
activities but as one of the most important factors in the development of my
work. I have stressed before how valuable I find the informality and
spontaneity of our discussion and the fact that one does not have to be on ones
guard when any issue is being argued. At the present time we have a group of
workers, each with some specialised knowledge and I believe that the free
interchange of ideas which has been so happily achieved and which, indeed,
was the basis for the founding of the Club, largely results from the fact that
questions of academic status do not arise." (Sholl 1952)
Young was the head of the Department of Anatomy at University College where Sholl
worked and although Sholl collaborated with him and continued to do so after the
club disbanded, in those days academic relations and the processes of career
advancement were such that he would have felt very uncomfortable with his boss as a
member. In any event the ‘no professors’ rule prevailed.
By the end of the summer of 1955 the club had run its course; many important
intellectual cross-fertilisations had occurred, and all had learnt much from each other.
In 1954 Turing had died in tragic and disturbing circumstances that have been well
documented (Hodges 1983). By now several member’s research had become very
well known internationally (e.g. Ashby and Walter in cybernetics, with Uttley not far
behind, and Rushton and Pringle in neurophysiology) and others were on the cusp of
major recognition. As careers advanced and families grew, many found it increasingly
difficult to justify the time needed for meetings. Another factor that may have played
a part in the club’s demise was that cybernetics had become respectable. Lord Adrian
had endorsed it in one of his Royal Society presidential addresses and talk of its
application in every conceivable branch of biology was rife. The frisson of antiestablishmentarianism that seeped into the early meetings was all but gone. The
September 1955 meeting, tacked on to the end of the London Symposium on
30
Information Theory, turned out to be the last. A reunion was held in November 1958
after Uttley’s Mechanization of Thought Processes symposium at the National
Physical Laboratory. Nine members turned up (Bates, Barlow, Dawson, Sholl, Slater,
Uttley, MacKay, Woodward, Hick); of the rest, Bates' note of the meeting reads:
"Absent: with expressed regret: Grey Walter, Merton, Westcott; with
expressed lack of interest: Ashby, McLardy; without expression: Rushton,
Pringle, Little, Good; emigrated: Gold, Shipton." (Bates 1958)
At the meeting, suggestions were put forward for possible new and younger members.
The first name recorded is Richard Gregory, then a young psychologist who had just
made his first professional presentation at the symposium; clearly, Bates had not lost
his ability to spot talent, as Gregory later became an extremely distinguished vision
scientist and Fellow of the Royal Society. However, the initiative came to nothing,
and the club did not meet again.
5. Themes
Although a very wide range of topics were discussed at club meetings, a number of
important themes dominated. These included: information theory, probabilistic and
statistical processes and techniques, pattern recognition, and digital versus analogue
models of the brain (Barlow 2002, 2007, Bates 1985). The themes usually surfaced in
the context of their application to understanding the nervous system and/or
developing machine intelligence.
5.1 Information theory
By far the greatest proportion of British war-time scientific effort had gone into radar
and communications, so it is perhaps unsurprising that there was huge interest in
information theory in the club. Many of the brain scientists realised very early on that
here was something that might be an important new tool in understanding the nervous
system. Shannon’s technical reports and papers were not easy to get hold of in Britain
in the late 1940s and so the first time Barlow came across them was when Bates sent
him copies – with a note to the effect that this was important stuff – along with his
invitation to join the club. Barlow agreed with Bates, immediately grasping the fact
that information theory provided a new, potentially measurable quantity that might
help to give a stronger theoretical underpinning to neurophysiology. Over the next
few years, as he learned more about the subject at club meetings – particularly from
Woodward – he developed a theoretical framework that shaped his research and
helped to propel him to the forefront of his field. Barlow used information theoretic
ideas in an implicit way in his now classic 1953 paper on the frog’s retina (Barlow
1953). This paper gives the first suggestion that the retina acts as a filter passing on
useful information, developing the idea that certain types of cells act as specialised
‘fly detectors’ – that the visual system has evolved to efficiently extract pertinent
information from the environment, an idea that was to become very influential. Later,
in a series of very important theoretical papers, he argued that the nervous system
may be transforming ‘sensory messages’ through a succession of recoding operations
31
which reduce redundancy in order to make the barrage of sensory information
reaching it manageable (Barlow 1959, 1961). (Reducing the amount of redundancy in
a message’s coding is one way to compress it and thereby make its transmission more
efficient.) This line of reasoning fed into the later development of his equally
influential ‘neuron doctrine for perceptual psychology’ which postulated that the brain
makes use of highly sparse neural ‘representations’ (Barlow 1972). As more
neurophysiological data became available, the notion of redundancy reduction became
difficult to sustain and Barlow began to argue for the principle of redundancy
exploitation in the nervous system. In work that has become very influential in
machine learning and computational neuroscience, Barlow and co-workers have
demonstrated how learning can be more efficient with increased redundancy as this
reduces ‘overlap’ between distributed patterns of activity (Gardner-Medwin and
Barlow 2001). For further discussion of these matters see the interview with Barlow
later in this volume (Barlow 2007).
Information and its role in biology was at the heart of many club debates. Mackay
believed the Shannon formulation was too restrictive and during the Ratio years he
developed his own set of ideas, allied with Gabor’s version of information theory
(Gabor 1946), that took account of context and meaning (Mackay 1952a,b). In the
early period of the club’s existence Ashby was working hard on the final version of
Design for a Brain and his habit of quizzing members on specific topics that would
help him refine the ideas in the book left several members with the impression that
that he was exclusively preoccupied with his own ideas and not open to new
influences. However, his journals indicate that he was becoming convinced of the
importance of information theory. He records a conversation with Gold and Pringle at
one meeting in 1950 on how much information was needed to specify a particular
machine, and by extension how much information must be encoded in the genes of an
animal. His arguments were demolished by Gold who pointed out that “complexity
doesn’t necessarily need any number of genes for its production: the most
complicated organisation can be produced as a result of a single bit of information
once the producing machinery has been set up” (Ashby 1950c). As usual Gold was
decades ahead in stressing the importance of genotype to phenotype mappings and the
role of development. This theme resurfaced in Ashby’s 1952 paper Can a mechanical
chess-player out play its designer (Ashby 1952b) in which he used information theory
to try and show how it might be possible to construct a machine whose behaviour
goes beyond the bounds of the specifications described by its designer. This paper
caused debate within the club, with Hick in particular disagreeing with Ashby’s claim
that random processes (such as mutations in evolution) can be a source of
information. This resulted in Hick joining in the discussion of Ashby’s paper on the
pages of The British Journal for the Philosophy of Science, where the original work
had appeared (Ashby 1952b), a debate which also included a contribution from J.B.S.
Haldane (Haldane 1952).
A striking example of the degree of enthusiasm for information theoretic ideas within
the club is given by the contents page of the first ever issue of the IEEE Transactions
on Information Theory, the field’s premier journal. This issue was based on the
proceedings of the 1st London Symposium on Information Theory, held in September
1950, and as such was completely dominated by Ratio Club members (IEEE
Transactions on Information Theory 1:1, Feb. 1953). Of the 22 full papers that were
32
published in it, 14 were from Club members. Of the remaining eight, three were by
Shannon and two by Gabor.
5.2 Probability and statistics
Probabilistic and statistical methods and processes were also of central concern to
many members in areas other than information. Good was a leading statistician who
pioneered various Bayesian ‘weight of evidence’ approaches (Good 1950b),
something that partly stemmed from his war-time code cracking work with Turing,
who also had a keen interest in the subject. Good led a number of club discussions
and debates on related topics which may have influenced Uttley’s ground-breaking
work on conditional probability machines for learning and reasoning (Uttley 1956). In
recent years similar approaches to those pioneered by these two have become very
prominent in machine learning. Woodward was very knowledgeable on probability
theory and gave, by request, at least one lecture to the club on the subject. Slater was
one of the first psychiatrists to use well grounded statistical techniques and did much
to try and make psychiatry, and medicine in general, more rigorously scientific.
Likewise, Sholl, whose first degree had been in statistics, introduced statistical
methods to the study of the anatomy of the nervous system. Many of the brain
scientists in the club were concerned with signal to noise problems in their practical
work and Barlow remembers that this was a regular topic of discussion (Barlow
2006). He recalls that Gold had deep and useful engineering intuitions on the subject.
As has been mentioned, Dawson was the leading expert on extracting clean EEG
signals in a clinical setting.
A related area that prompted much discussion was that of the possible roles of random
processes and structures in the nervous system. It has already been noted that Pringle
and Ashby gave presentations in this area, but Barlow remembers that many other
members, including Turing, were intrigued by the topic (Barlow 2002).
5.3 Philosophy
A quick glance at the meeting titles shown in Table 1, and the topics of introductory
talks listed in Section 4, make it obvious that many club discussions had a distinctly
philosophical flavour. Mackay was particularly keen to turn the conversation in that
direction, prompting Hodges to refer to him as ‘a philosophical physicist’ in a
mention of a Ratio meeting in his biography of Turing (Hodges 1983 p.411), and
Woodward recalls that it was a good idea to keep him off the subject of Wittgenstein!
(Woodward 2002)
5.4 Pattern recognition
Pattern recognition was another hot topic, in relation to both natural and machine
intelligence. The ninth meeting of the club on the 18th May 1950 was dedicated to this
subject, then very much in its infancy; the perspectives of a number of members were
followed by a general free-for-all discussion. Ashby provided a hand-out in which he
tried to define ‘recognition’ and ‘pattern’, concluding that a large part of pattern
recognition is classification or categorisation, and wondering “can class-recognition
profitably be treated as a dissection of the total information into two parts – a part that
33
identifies the inputs’ class, and a part that identifies the details within the class?”
(Ashby 1950d). Grey Walter also provided a hand-out – a set of condensed and hasty
notes in which he concentrated on a brief survey of types of pattern recognition
problems and techniques. He noted that “recognition of pattern correlates well with
‘intelligence’; only highest wits can detect patterns in top Raven Matrices where the
symmetry is abstract not graphic. Likewise in ‘good’ music, odours (not so much in
man).” (Walter 1950b) We can be sure that a vigorous debate ensued!
There is no room to discuss the other equally interesting themes that arose in club
discussions, such as motor control mechanisms in humans and animals (Merton and
Pringle were particularly expert in this area, with Westcott providing the engineering
perspective), analogue versus digital models of the functioning of the nervous system
(see the interview with Barlow later in this book for discussion of this in relation to
the Ratio Club) and the relationship between evolution and learning (Pringle wrote an
important paper on this at the time of the club (Pringle 1951) which, as Cowan has
pointed out (Cowan 2003), laid the foundations for what later became known as
competitive learning).
5.5 Artefacts and the synthetic method
However, there is one last implicit theme that is important enough to deserve some
brief discussion: the use of artefacts within the synthetic method. In addition to the
engineers, several other members were adept at designing and constructing
experimental equipment (often built from surplus military components left over from
the war). This tendency was naturally transferred to an approach referred to by Craik
as the ‘synthetic method’ – the use of physical models to test and probe neurological
or psychological hypotheses. In this spirit Ashby and Walter developed devices that
were to become the most famous of all cybernetic machines: Ashby’s Homeostat and
Walter’s Tortoises. Both machines made headlines around the world, in particular the
tortoises which featured in newsreels and television broadcasts, and were exhibited at
the Festival of Britain (Holland 2003).
The Homeostat was an electromechanical device intended to demonstrate Ashby’s
theory of ultrastable systems – adaptive systems making use of a double feedback
mechanisms in order to keep certain significant quantities within permissible ranges.
As mentioned earlier, these essential variables represented such things as blood
pressure or body temperature in an animal. According to Ashby, ultrastable systems
were at the heart of the generation of adaptive behaviour in biological systems. Part of
the device is shown in Figure 3.
The machine consisted of four units. On top of each was a pivoted magnet. The
angular deviation of the four magnets represented the main variables of the system.
The units were joined together so that each sent its output to the other three. The
torque on each magnet was proportional to the total input current to the unit. The units
were constructed such that their output was proportional to the deviation of their
magnet from the central position. The values of various commutators and
potentiometers acted as parameters to the system: they determined its subsequent
behaviour. The electrical interactions between the units modelled the primary
feedback mechanisms of an ultrastable system. A secondary feedback mechanism was
implemented via switching circuitry to make pseudo-random (step) changes to the
34
parameters of the system by changing potentiometer and commutator values. This
mechanism was triggered when one of the essential variables (proportional to the
magnet’s deviation) went out of bounds. The system continued to reset parameters
until a stable configuration was reached whereby no essential variable were out of
range and the secondary feedback mechanisms became inoperative. The units could
Figure 3: The Homeostat. Two of the four units can be seen.
be viewed as abstract representations of an organism interacting with its environment.
Ultrastability was demonstrated by first taking control of one of the units by reversing
the commutator by hand, thereby causing an instability, and then observing how the
system adapted its configuration until it found a stable state once more (for full details
see Ashby 1952a).
On November the 20th 1946 Turing had written to Ashby after being passed a letter
from Ashby to Sir Charles Darwin, director of the National Physical Laboratory,
distinguished mathematician and grandson of the Charles Darwin (and therefore
Horace Barlow’s uncle). Ashby had enquired about the future suitability of the
planned ACE digital computer, which was being designed at the NPL by Turing and
others, for modelling brain-like mechanisms. We can assume he was thinking of the
35
possibility of using the computer to develop a programmed equivalent of what was to
become his famous Homeostat. In his reply, Turing enthusiastically endorsed such an
idea, telling Ashby that “..in working on the ACE I am more interested in the
possibility of producing models of the action of the brain than in the practical
applications of computing.” (Turing 1946). Turing explained that in theory it would
be possible to use the ACE to model adaptive processes by making use of the fact that
it would be, in all reasonable cases, a universal machine. He went on to suggest that
“.. you would be well advised to take advantage of this principle, and do your
experiments on the ACE, instead of building a special machine. I should be very glad
to help you over this.” Unfortunately this collaboration never materialised. Turing
withdrew from the ACE project following the NPL management’s inability, or
unwillingness, to properly mange the construction of the machine (Hodges, 1983).
Although the ACE project stalled, Ashby’s notebooks from 1948 show that he was
still musing over the possibility of using a computer to demonstrate his theories and
was able to convinces himself that the ACE could do the job (Ashby 1948a). A pilot
ACE digital computer was finally finished in mid 1950, but in the meantime a
physical Homeostat had been finished in 1948 (Ashby 1948b). The Manchester
Mark1, often regarded as the world’s first full-scale stored-programme digital
computer and the project with which Turing was by then associated, was finished a
few months after this. It is very interesting to note that Ashby was considering using a
general purpose programmable digital computer to demonstrate and explore his
theories before any such machine even existed. It would be many years before
computational modelling became commonplace in science.
Grey Walter's tortoises were probably the first ever wheeled mobile autonomous
robots. The devices were three-wheeled and turtle-like, sporting a protective ‘shell’
(see Figure 4). These vehicles had a light sensor, touch sensor, propulsion motor,
steering motor, and an electronic valve based analogue ‘nervous system’. Walter’s
intention was to show that even in a very simple nervous system (the tortoises had
two artificial neurons) complexity could arise out of the interactions between its units.
By studying whole embodied sensorimotor systems, he was pioneering a style of
research that was to become very prominent in AI many years later, and remains so
today (Brooks 1999, PRS 2003). Between Easter 1948 and Christmas 1949, he built
the first tortoises, Elmer and Elsie. They had similar circuits and electronics, but their
shells and motors were a little different. They were rather unreliable and required
frequent attention. . The robots were capable of phototaxis, by which they could find
their way to a recharging station when they ran low on battery power. In 1951, his
technician, Mr. W.J. 'Bunny' Warren, designed and built six new tortoises for him to a
high professional standard. Three of these tortoises were exhibited at the Festival of
Britain in 1951; others were demonstrated in public regularly throughout the fifties.
He referred to the devices as Machina Speculatrix after their apparent tendency to
speculatively explore their environment.
Walter was able to demonstrate a variety of interesting behaviours as the robots
interacted with their environment and each other (Walter 1950a, 1953). In one
experiment he watched as the robot moved in front of a mirror, responding to its own
reflection. "It began flickering," he wrote. "Twittering, and jigging like a clumsy
Narcissus." Walter argued that if this behaviour was observed in an animal it "might
be accepted as evidence of some degree of self-awareness."
36
One or other of the machines was demonstrated at at least one Ratio meeting. Tommy
Gold recalled being fascinated by it and wondering if the kind of principle underlying
its behaviour could be adapted to develop autonomous lawnmowers (Gold 2002),
something that came to pass many decades later.
There was much discussion in meetings of what kind of intelligent behaviour might
be possible in artefacts and, more specifically, how the new general purpose
computers might exhibit mind-like behaviour. Mackay was quick to point out that
“the comparison of contemporary calculating machines with human brains appears to
have little merit, and has done much to befog the real issue, as to how far an artefact
could in principle be made to show behaviour of the type which we normally regard
as characteristic of a human mind.” (Mackay 1951)
Figure 4: Grey Walter watches one of his tortoises push aside some wooden blocks on its way
back to its recharging hutch. Circa 1952.
6. Interdisciplinarity
From what we have seen of its founding and membership, to comment that the Ratio
Club was an interdisciplinary organization is stating the obvious. However, what is
37
interesting is that is was a successful interdisciplinary venture. This was partly a
function of the time, when recent war-time work and experiences encouraged the
breaking down of barriers, and partly a function of Bates’ keen eye for the right
people. Even when war work was factored out, many of the members had very broad
backgrounds. To give a few examples: Sholl had moved from mathematical sciences
to anatomy following earlier studies in theology, zoology and physiology, Uttley had
degrees in mathematics and psychology, Merton was a brilliant natural engineer (he
and Dawson were later instrumental in the adoption of digital computing techniques
in experimental neurophysiology). All the brain scientists had strong interests, usually
going back many years, in the use of mathematical and quantitative techniques. There
was a similar, if less marked, story among the engineers and mathematicians: we have
already commented on Gold’s disregard for disciplinary boundaries, Turing was
working on biological modelling and Mackay had started his conversion into a
neuropsychologist. Most members were open-minded, with wide-ranging interests
outside science. This mix allowed important issues to be discussed from genuinely
different perspectives, sparking off new insight.
Most members carried this spirit with them throughout their careers and many were
involved in an extraordinarily wide range of research, even if this was within a single
field. This lack of narrowness meant that most had other string to their bows (several
were very good musicians and a number were involved with other areas of the arts),
sometimes starting whole new careers in retirement (Woodward’s enormous success
in clock making has been mentioned, and in later life Slater became an expert on the
use of statistical evidence in analysing the authorship of Shakespearean texts).
Figure 6: Philip Woodward at home in 2002. In
the background is one of the mechanical clocks he
has designed and built. His W5 clock is one of the
most accurate pendulum controlled clocks ever
made. It has won a number of international
awards and helped to make Woodward one of the
most celebrated horologists of our times.
Figure 5: Jack Good at home in 2002. The
sculpture above his head, ‘Jack Good’s
Dream’, was made in glass by an artist
friend and exhibited at the famous 1968
Cybernetic Serendipity show at the ICA,
London. It is based on a geometric
construction of intersecting cylinders and
spheres – the formation came to Jack in a
dream.
38
A key ingredient in the club’s success was its informal, relaxed character, which
encouraged unconstrained contributions and made meetings fun, and the fact that it
had a fairly strong focus right from the start: new ways of looking at mechanisms
underlying intelligent behaviour, particularly from a biological perspective.
7. The legacy of the club
In the USA the cybernetics movement organised the Macy Foundation conferences
whose published proceedings were made available a year or so after each meeting;
they consisted of verbatim transcripts, lightly edited by Heinz von Foerster, and so the
substance of all the presentations and discussions was readily available to the
academic community and the public, where they had considerable influence. In
contrast, no detailed records of the Ratio Club's meetings were made, let alone
circulated or published, and so, in assessing the influence of the Ratio Club, it is clear
that it can only have been of two kinds: the influence of its members on one another,
and the consequences of that influence for their own work.
Unravelling such influences is non-trivial, but we have already seen testaments from
several members on how important the club was to the development of their research.
In 1981, after coming across some long forgotten Ratio material, Pringle was
prompted to write to Bates:
“Dear John,
Going through some drawers of papers today in the lab, I came across a
photograph of 17 members of the Ratio Club … It occurs to me that someone
ought to write up the history of the club, since it was in the old 17 th century
tradition and, to me at any rate, was a most valuable stimulus at a time when I
was only just getting back into biology after the war.” (Pringle 1981)
He also wrote to Mackay, who agreed on the importance of the club and sent his Ratio
papers to help with the history Pringle and Bates planned to put together.
Unfortunately this venture stalled.
Pringle’s response to the club was typical of its effect on many members, particularly
the biologists: it acted as an inspiration and a spur. Much subsequent work of
members had its origins, at least partially, in club discussions. The important
influence on Barlow has already been explained; given his major impact on
neuroscience, if all the club had done was to put Barlow on the road he travelled, it
would be of significance. Clearly it did much more than that. As a mark of his debt to
the Ratio Club, Uttley included the photograph of it shown in Figure 2 in his 1979
book, Information Transmission in the Nervous System (Uttley 1979). The influence
of the biologist in the club appears to have played an important role in Mackay’s
transformation from physicist to prominent neuropsychologist. The pages of Ashby’s
private journals, in which he meticulously recorded his scientific ideas as they
developed, show that the club had some influence on him, although how much is hard
39
to judge – before becoming very well known, he had worked on his theories in
isolation for years, and there was always something of the outsider about him. His
grandson, John, has pointed out that Ashby’s most prolific years, as far as scientific
journal writing was concerned, exactly coincide with the Ratio years (Ashby, J. 2004).
In all events he was an active member who rarely missed a meeting.
Most members went on to pursue highly distinguished careers. Many gained
professorships at prestigious universities. Between them club members were awarded
a host of prizes and honours, including seven Fellowships of the Royal Society and a
CBE to Slater for services to psychiatry. Four members (Barlow, Rushton, Gold,
Walter) came within striking distance of a Nobel Prize (many feel that at least
Rushton and Barlow should have received one) and Turing’s work is likely to be
remembered for centuries. Many papers and books written by members of the group,
including those produced during the Ratio years, are still widely cited, with many
ideas and techniques emanating from members very much in currency today.
Uttley and Mackay went on to set up and run successful interdisciplinary groups, at
the NPL and Keele University respectively; it is likely that their experience of the
extraordinary club influenced them in these ventures.
So how should we assess the club's contribution? It seems to have served a number of
purposes during a narrow and very specific window in time. It influenced a relatively
small group of British scientists in their post-war careers; but given the degree of
eminence many of them reached, and their influence on subsequent generations, this
turned out to be highly significant. It certainly concentrated and channeled the
cybernetic currents that had developed independently in the UK during the war. It also
provided a conduit for the new ideas from the US to be integrated into work in the
UK. It stimulated the introduction into biology of cybernetic ideas, and in particular
of information theory. And, perhaps appropriately for a cybernetic organisation, it
stopped meeting when these purposes had been achieved.
Despite its length, this paper can only act as an introduction to the life and times of
the club and its members. We were forced to leave out a huge amount of fascinating
material; there is still much to tell.
Acknowledgements: We owe a great debt of gratitude to the surviving members of
the Ratio Club who all generously participated in the research for this article – Horace
Barlow, Jack Good, Harold Shipton (who died as this book went to press), John
Westcott and Philip Woodward, and to the late Tommy Gold who we interviewed
before his death. Papers provided by John Westcott and Jack Good have been
enormously helpful. Thanks also to the many people who helped with background
information and other material, in particular John and Mick Ashby, Jack Cowan,
Richard Gregory, Peter Asaro, Igor Alexander, Helen Morton, Michael Slater, Ann
Pasternak Slater, the late John Maynard Smith, the late Dick Grimsdale, Janet Shipton
and Andrew Hodges. Thanks to the following for very useful discussions of this and
related material: Jon Bird, Peter Cariani, Maggie Boden, Roland Baddeley, Danny
Osorio and Emmet Spier.
References
40
Ashby, J. (2004) Talk at W.Ross Ashby Centenary Conference, University of Illinois.
Ashby, W.R. (1940) Adaptiveness and equilibrium. J. Ment. Sci. 86, 478.
Ashby, W.R. (1944) Letter to Kenneth Craik, 6th June 1944. W.Ross Ashby Archive,
British Library, London.
Ashby, W.R. (1946) Dynamics of the Cerebral Cortex: The Behavioral Properties of
Systems in Equilibrium, American Journal of Psychology 59(4), 682-686.
Ashby, W.R. (1947) Letter to W.E. Hick, 14th July 1947. W.Ross Ashby Archive,
British Library, London.
Ashby, W.R. (1948a) Ashby’s journal, 1928-1972, pp.2520-2523, 3rd June 1948, W.
Ross Ashby’s Digital Archive, http://www.rossashby.info/journal
Ashby, W.R. (1948b) Design for a Brain. Electronic Engineering, 20:379-383.
Ashby, W.R. (1949a) Ashby’s journal, 1928-1972, p.2624, W. Ross Ashby’s Digital
Archive, http://www.rossashby.info/journal
Ashby, W.R. (1949b) Review of Wiener’s Cybernetics, Journal of Mental Science,
95, 716-724.
Ashby, W.R. (1950a) Subjects for discussion, 18th Feb. 1950. Unpublished papers
and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Ashby, W.R. (1950b) Suggested topics for discussion, 15th May 1950. Unpublished
Ratio Club papers of John Westcott.
Ashby, W.R. (1950c) Ashby’s journal, 1928-1972, p.2806, (28th April 1950) W. Ross
Ashby’s Digital Archive, http://www.rossashby.info/journal
Ashby, W.R. (1950d) Pattern recognition in animals and machines, short paper for
Ratio Club, May 1950. Unpublished Ratio Club papers of John Westcott.
Ashby, W.R. (1952a) Design for a Brain. Chapman and Hall.
Ashby, W.R. (1952b) Can a mechanical chess-player outplay its designer? The British
Journal for the Philosophy of Science, 3(9):44-57.
Ashby, W.R. (1958) An Introduction to Cybernetics. Chapman and Hall.
Barlow, H.B. (1949). Letter to John Bates, August 1949, Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Barlow, H.B. (1953). Summation and inhibition in the frog’s retina. Journal of
Physiology, 119:69-88.
41
Barlow, H.B. (1959) Sensory mechanism, the reduction of redundancy, and
intelligence. In A. Uttley (Ed.) Mechanisation of Thought Processes : Proceedings of
a Symposium held at the National Physical Laboratory on 24-27 November 1958. Her
Majesty's Stationery Office, London, 537—559.
Barlow, H.B. (1961) Possible principles underlying the transformations of sensory
messages. In W. A. Rosenblith (Ed), Sensory Communication, MIT Press, 217-234.
Barlow, H. B. (1972). Single units and sensation: A neuron doctrine for perceptual
psychology? Perception 1:371–394.
Barlow, H.B. (1986) William Rushton. Biographical Memoirs of Fellows of the Royal
Society, 32:423-459.
Barlow, H.B. (2001) Interview with PH, 30th March 2001, Cambridge.
Barlow, H.B. (2002) Interview with PH and OH, 19th June 2002, Cambridge.
Barlow, H.B. (2007) An Interview with Horace Barlow. In P. Husbands, O. Holland,
M. Wheeler (Eds), The Mechanisation of Mind in History, MIT Press (this volume).
Bates, J. (1945) Letter to W.E. Hick, 30th May 1945, J.A.V. Bates Archive, The
Wellcome Library for the History and Understanding of Medicine, London.
Bates, J. (1949a) Letter to Grey Walter, 27th July 1949, Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1949b) Letter to John Westcott, 3rd August 1949, Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1949c) Letter to William Hick, 17th August 1949, Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1949d) Initial membership list of Ratio Club, 14th September 1949. At the
bottom of the list Bates reproduced ‘extracts from a Latin dictionary’, submitted by
Uttley, giving the definition of ratio and various other words using it as a root.
Unpublished papers and records for the Ratio Club. J.A.V. Bates Archive, The
Wellcome Library for the History and Understanding of Medicine, London.
Bates, J. (1949e) Notes for first meeting of the Ratio Club, 14th September 1949.
Unpublished papers and records for the Ratio Club. J.A.V. Bates Archive, The
Wellcome Library for the History and Understanding of Medicine, London.
Bates, J. (1949f) Letter to Grey Walter, 4th October 1949. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
42
Bates, J. (1949g) Memo to Ratio Club members, (undated, but between first and
second meeting, probably late Sept. 1949). Unpublished Ratio Club papers of John
Westcott.
Bates, J. (1950a) Letter to Jack Good, 13th December 1950. Unpublished papers of
I.J. Good.
Bates, J. (1950b) Letter to Grey Walter, 9th December 1950. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1951) The Ratio Club, memo to members, December 1951. Unpublished
papers and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library
for the History and Understanding of Medicine, London.
Bates, J. (1952a) Membership list of the Ratio Club, Jan. 1952, Unpublished papers
and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1952b) Draft paper, J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Bates, J. (1958) Note on final meeting, Unpublished papers and records for the Ratio
Club. J.A.V. Bates Archive, The Wellcome Library for the History and Understanding
of Medicine, London.
Bates, J. (1985) Notes for an article on the Ratio Club. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
BBC History (2005) WW2 People’s war (www.bbc.co.uk/ww2peopleswar/) Article
ID: A6038453.
Blake, D. and Uttley, A. (Eds.) (1959) The Mechanisation of Thought Processes.
Volume 10 of National Physical Laboratory Symposia. Her Majesty's Stationary
Office, London.
Boden, M. A. (2006) Mind as Machine: A History of Cognitive Science. Oxford:
Oxford University Press.
Brindley, G. (2002) Interview with PH and OH, May 2002, London.
Brooks, R.A. (1999) Cambrian Intelligence: The Early History of the New AI, MIT
Press.
Churchill, W. (1949) Chapter 4: The wizard war, in The second world war, volume 2:
Their finest hour, Cassell and co., London.
43
Clark, D. (2003) Enclosing the field, PhD thesis, Department of Computer Science,
Warwick University.
Cowan, J. (2003) Interview with PH and OH, 6th April 2003,Chicago.
Craik, K.J.W. (1943) The Nature of Explanation, CUP.
Craik, K.J.W. (1944) Letter to W. Ross Ashby, 1st July 1944. W.Ross Ashby Archive,
British Library, London.
Craik, K.J.W. (1948). Theory of the human operator in control systems. II - Man as an
element in a control system. Brit. J. Psychol. 38, 142-148
Dawson, G. D. and Walter, W. G. (1944) The scope and limitation of visual and
automatic analysis of the E.E.G. J. Neurol. Neurosurg. Psychiat. 7, 119-130.
Dawson, G. D. (1954) A summation technique for the. detection of small evoked
potentials. Electroenceph. Clin. Neurophysiol. 6, 65-84.
Fleck, J. (1982) Development and establishment in Artificial Intelligence. In N. Elias
et al. (Eds), Scientific establishments and hierachies, Sociology of the Sciences, vol
VI, 169-217, D. Reidel Publishing Company.
Gabor, D. (1946) Theory of communication. J. IEE, 93(3):429-457.
Gardner, M.R. and Ashby, W.R. (1970) Connectance of large dynamic (cybernetic)
systems: Critical values for stability. Nature, 228:784.
Gardner-Medwin, A.R. and Barlow, H.B. (2001), The limits of counting accuracy in
distributed neural representations, Neural Computation 13(3):477-504.
Gold, T. (1948) Hearing. II. The Physical Basis of the Action of the Cochlea.
Proceedings of the Royal Society of London, Series B, 135, No. 881, 492-498.
Gold, T. (2002) Interview with PH and OH, Ithica, 17 th April 2002.
Good, I.J. (1950a) Letter to John Bates, 9th Dec. 1950. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Good, I.J. (1950b) Probability and the Weighing of Evidence. London: Charles
Griffin.
Haldane, J.B.S. (1952) The mechanical chess player. The British Journal for the
Philosophy of Science, 3(10):189-191.
Heims, S. (1991) Constructing a Social Science for Postwar America: The
Cybernetics Group, 1946-1953. MIT Press.
44
Hick, W.E. (1947a) Letter to W. Ross Ashby, 10th June 1947. W.Ross Ashby Archive,
British Library, London.
Hick, W.E. (1947b) Letter to W. Ross Ashby, 30th June 1947. Quoted in Ashby 1947.
Hick, W.E. (1952) On the rate of gain of information. Quarterly Journal of
Experimental Psychology, 4:11-26.
Hodges, A. (1983) Alan Turing: The enigma of intelligence, Counterpoint, London.
Holland, O. (2003) Exploration and high adventure: the legacy of Grey Walter. Phil.
Trans. R. Soc. Lond. A 361, 2085-2121.
Husbands, P. and Holland, O. (forthcoming) The Ratio Club.
Mackay, D.M. (1949). Letter to John Bates, 1st September 1949, Unpublished papers
and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Mackay, D.M. (1951) Mindlike Behaviour in Artefacts. The British Journal for the
Philosophy of Science, 2(6):105-121.
Mackay, D.M. (1952a) In search of basic symbols. In H. von Foerster (ed.),
Proceedings of the 8th conference on cybernetics, 1951, 181-221, New York: Josiah
Macey Jr. Foundation.
Mackay, D.M. (1952b) The nomenclature of information theory. In H. von Foerster
(ed.), Proceedings of the 8th conference on cybernetics, 1951, 222-235. New York:
Josiah Macey Jr. Foundation.
MacKay, D.M. (1991) Behind the Eye, Oxford: Blackwells.
McCarthy, J., Minsky, M., Rochester, N. and Shannon, C. (1955) A Proposal for the
Dartmouth Summer Research Project on Artificial Intelligence, wwwformal.stanford.edu/jmc/history/dartmouth/dartmouth.html
Merton, P.A. (1953) Slowly conducting muscle spindle afferents. Acta Physiol Scand.
29(1):87–88.
Merton, P.A. and Morton, H.B. (1980) Stimulation of the cerebral cortex in the intact
human subject. Nature 285:227.
Pringle, J.W. (1938) Proprioception in insects. I. A new type of mechanical receptor
from the palps of the cockroach. J. Exp. Biol. 15: 101-113.
Pringle, J.W. (1949) Letter to John Bates, 5th August 1949, Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
45
Pringle, J.W. (1951) On the parallel between learning and evolution. Behaviour 3:
174-215.
Pringle, J.W. (1981) Letter to John Bates, 13th January 1981. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
PRS (2003) Phil. Trans. Royal Society Lond. A 361 (special issue on the legacy of W.
Grey Walter).
Rosenblueth, A., Wiener N. and Bigelo, J. (1943). Behaviour, purpose and teleology.
Phil. Sci. 10(1), 18-24.
Rushton, W. (1935) A theory of excitation. J. Physiol. 84:42.
Rushton, W. (1955) Foveal photopigments in normal and colour-blind. J. Physiol.
129:41-42.
Shannon, C. and Weaver, W. (1949) The Mathematical Theory of Communication,
University of Illinois Press.
Shipton H&J (2002) Interview of Harold and Janet Shipton by PH and OH, Jupiter,
FL, October 2002.
Sholl, D.A. (1952) Letter to John Bates, 28 May 1952. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Sholl, D.A. (1956) The Organization of the Nervous System, McGraw-Hill.
Slater, E., Gottesman, I., and Shields, J. (1971) Man, Mind, and Heredity: Selected
Papers of Eliot Slater on Psychiatry and Genetics, Johns Hopkins Press.
Tattersall, M. (2006). Personal communication based on information in his father’s
diary. His father was a fellow medical officer and POW with McLardy.
Turing, A.M. (1936) On computable numbers, with an application to the
Entscheidungsproblem, Proc. London Math. Soc. (2), 42: 230-265.
Turing, A.M. (1946) Letter to W. Ross Ashby, undated (about 19 Nov. 1946).
W.Ross Ashby Archive, British Library, London.
Turing, A.M. (1950) Computing machinery and intelligence, Mind, 59, 433460.
Turing, A.M. (1952). The chemical basis of morphogenesis, Phil. Trans. R. Soc.
London B. 237:37-72.
46
Uttley, A.M. (1956) Conditional Probability Machines and Conditioned Reflexes. In
C.E. Shannon & J. McCarthy (Eds.), Automata Studies, Princeton: Princeton
University Press, 253-275.
Uttley, A.M. (1979) Information Transmission in the Nervous System, Academic
Press.
Uttley, A.M. (1982) Brain, Mind and Spirit, privately published.
Walter, W. G. (1947) Letter from Grey Walter to Professor Adrian, 12 June 1947.
Burden Neurological Institute Papers, Science Museum, London.
Walter, W. G. (1949) Letter to John Bates, 29th September 1949. Unpublished papers
and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Walter, W.G. (1950a) An imitation of life, Scientific American 182(5): 42-45.
Walter, W.G. (1950b) Pattern Recognition, short paper for Ratio Club, 15 th May
1950. Unpublished Ratio Club papers of John Westcott.
Walter, W.G. (1951) Letter to John Bates, November 1951. Unpublished papers and
records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Walter, W.G. and Shipton, H. (1951) A new toposcopic display system.
Electroenceph. Clin. Neurophysiol., 3:281-292.
Walter, W. G. (1953) The Living Brain. London: Duckworth.
Westcott, J. (1949-53) Unpublished notebook used at Ratio Club meetings 19491953.
Westcott, J. (2002) Interview with PH and OH, Imperial College, London, 15 th March
2002.
Wiener, N. (1948) Cybernetics, or control and communication in the Animal and the
Machine. MIT Press.
Wisdom, J.O. (1951) The Hypothesis of Cybernetics, The British Journal for the
Philosophy of Science, 2(5), 1-24.
Woodward, P.M. (1951) Letter to John Bates, 29th October 1951. Unpublished papers
and records for the Ratio Club. J.A.V. Bates Archive, The Wellcome Library for the
History and Understanding of Medicine, London.
Woodward, P.M. (1953) Probability and Information Theory, with applications to
radar. Pergamon Press.
47
Woodward, P.M. (1995) My Own Right Time: An exploration of clockwork design.
OUP.
Woodward, P.M. (2002) Interview with PH and OH, Malvern, 8 th May 2002.
48